WorldWideScience

Sample records for convolution model obmennaya

  1. Model structure selection in convolutive mixtures

    DEFF Research Database (Denmark)

    Dyrholm, Mads; Makeig, S.; Hansen, Lars Kai

    2006-01-01

    The CICAAR algorithm (convolutive independent component analysis with an auto-regressive inverse model) allows separation of white (i.i.d) source signals from convolutive mixtures. We introduce a source color model as a simple extension to the CICAAR which allows for a more parsimonious represent......The CICAAR algorithm (convolutive independent component analysis with an auto-regressive inverse model) allows separation of white (i.i.d) source signals from convolutive mixtures. We introduce a source color model as a simple extension to the CICAAR which allows for a more parsimonious...... representation in many practical mixtures. The new filter-CICAAR allows Bayesian model selection and can help answer questions like: ’Are we actually dealing with a convolutive mixture?’. We try to answer this question for EEG data....

  2. Incomplete convolutions in production and inventory models

    NARCIS (Netherlands)

    Houtum, van G.J.J.A.N.; Zijm, W.H.M.

    1997-01-01

    In this paper, we study incomplete convolutions of continuous distribution functions, as they appear in the analysis of (multi-stage) production and inventory systems. Three example systems are discussed where these incomplete convolutions naturally arise. We derive explicit, nonrecursive formulae

  3. Spherical convolutions and their application in molecular modelling

    DEFF Research Database (Denmark)

    Boomsma, Wouter; Frellsen, Jes

    2017-01-01

    Convolutional neural networks are increasingly used outside the domain of image analysis, in particular in various areas of the natural sciences concerned with spatial data. Such networks often work out-of-the box, and in some cases entire model architectures from image analysis can be carried over...... to other problem domains almost unaltered. Unfortunately, this convenience does not trivially extend to data in non-euclidean spaces, such as spherical data. In this paper, we introduce two strategies for conducting convolutions on the sphere, using either a spherical-polar grid or a grid based...... of spherical convolutions in the context of molecular modelling, by considering structural environments within proteins. We show that the models are capable of learning non-trivial functions in these molecular environments, and that our spherical convolutions generally outperform standard 3D convolutions...

  4. CICAAR - Convolutive ICA with an Auto-Regressive Inverse Model

    DEFF Research Database (Denmark)

    Dyrholm, Mads; Hansen, Lars Kai

    2004-01-01

    We invoke an auto-regressive IIR inverse model for convolutive ICA and derive expressions for the likelihood and its gradient. We argue that optimization will give a stable inverse. When there are more sensors than sources the mixing model parameters are estimated in a second step by least square...... estimation. We demonstrate the method on synthetic data and finally separate speech and music in a real room recording....

  5. Edgeworth Expansion Based Model for the Convolutional Noise pdf

    Directory of Open Access Journals (Sweden)

    Yonatan Rivlin

    2014-01-01

    Full Text Available Recently, the Edgeworth expansion up to order 4 was used to represent the convolutional noise probability density function (pdf in the conditional expectation calculations where the source pdf was modeled with the maximum entropy density approximation technique. However, the applied Lagrange multipliers were not the appropriate ones for the chosen model for the convolutional noise pdf. In this paper we use the Edgeworth expansion up to order 4 and up to order 6 to model the convolutional noise pdf. We derive the appropriate Lagrange multipliers, thus obtaining new closed-form approximated expressions for the conditional expectation and mean square error (MSE as a byproduct. Simulation results indicate hardly any equalization improvement with Edgeworth expansion up to order 4 when using optimal Lagrange multipliers over a nonoptimal set. In addition, there is no justification for using the Edgeworth expansion up to order 6 over the Edgeworth expansion up to order 4 for the 16QAM and easy channel case. However, Edgeworth expansion up to order 6 leads to improved equalization performance compared to the Edgeworth expansion up to order 4 for the 16QAM and hard channel case as well as for the case where the 64QAM is sent via an easy channel.

  6. Accurate lithography simulation model based on convolutional neural networks

    Science.gov (United States)

    Watanabe, Yuki; Kimura, Taiki; Matsunawa, Tetsuaki; Nojima, Shigeki

    2017-07-01

    Lithography simulation is an essential technique for today's semiconductor manufacturing process. In order to calculate an entire chip in realistic time, compact resist model is commonly used. The model is established for faster calculation. To have accurate compact resist model, it is necessary to fix a complicated non-linear model function. However, it is difficult to decide an appropriate function manually because there are many options. This paper proposes a new compact resist model using CNN (Convolutional Neural Networks) which is one of deep learning techniques. CNN model makes it possible to determine an appropriate model function and achieve accurate simulation. Experimental results show CNN model can reduce CD prediction errors by 70% compared with the conventional model.

  7. Model selection for convolutive ICA with an application to spatiotemporal analysis of EEG

    DEFF Research Database (Denmark)

    Dyrholm, Mads; Makeig, S.; Hansen, Lars Kai

    2007-01-01

    We present a new algorithm for maximum likelihood convolutive independent component analysis (ICA) in which components are unmixed using stable autoregressive filters determined implicitly by estimating a convolutive model of the mixing process. By introducing a convolutive mixing model...... for the components, we show how the order of the filters in the model can be correctly detected using Bayesian model selection. We demonstrate a framework for deconvolving a subspace of independent components in electroencephalography (EEG). Initial results suggest that in some cases, convolutive mixing may...

  8. An effective convolutional neural network model for Chinese sentiment analysis

    Science.gov (United States)

    Zhang, Yu; Chen, Mengdong; Liu, Lianzhong; Wang, Yadong

    2017-06-01

    Nowadays microblog is getting more and more popular. People are increasingly accustomed to expressing their opinions on Twitter, Facebook and Sina Weibo. Sentiment analysis of microblog has received significant attention, both in academia and in industry. So far, Chinese microblog exploration still needs lots of further work. In recent years CNN has also been used to deal with NLP tasks, and already achieved good results. However, these methods ignore the effective use of a large number of existing sentimental resources. For this purpose, we propose a Lexicon-based Sentiment Convolutional Neural Networks (LSCNN) model focus on Weibo's sentiment analysis, which combines two CNNs, trained individually base on sentiment features and word embedding, at the fully connected hidden layer. The experimental results show that our model outperforms the CNN model only with word embedding features on microblog sentiment analysis task.

  9. A staggered-grid convolutional differentiator for elastic wave modelling

    Science.gov (United States)

    Sun, Weijia; Zhou, Binzhong; Fu, Li-Yun

    2015-11-01

    The computation of derivatives in governing partial differential equations is one of the most investigated subjects in the numerical simulation of physical wave propagation. An analytical staggered-grid convolutional differentiator (CD) for first-order velocity-stress elastic wave equations is derived in this paper by inverse Fourier transformation of the band-limited spectrum of a first derivative operator. A taper window function is used to truncate the infinite staggered-grid CD stencil. The truncated CD operator is almost as accurate as the analytical solution, and as efficient as the finite-difference (FD) method. The selection of window functions will influence the accuracy of the CD operator in wave simulation. We search for the optimal Gaussian windows for different order CDs by minimizing the spectral error of the derivative and comparing the windows with the normal Hanning window function for tapering the CD operators. It is found that the optimal Gaussian window appears to be similar to the Hanning window function for tapering the same CD operator. We investigate the accuracy of the windowed CD operator and the staggered-grid FD method with different orders. Compared to the conventional staggered-grid FD method, a short staggered-grid CD operator achieves an accuracy equivalent to that of a long FD operator, with lower computational costs. For example, an 8th order staggered-grid CD operator can achieve the same accuracy of a 16th order staggered-grid FD algorithm but with half of the computational resources and time required. Numerical examples from a homogeneous model and a crustal waveguide model are used to illustrate the superiority of the CD operators over the conventional staggered-grid FD operators for the simulation of wave propagations.

  10. A deep convolutional neural network model to classify heartbeats.

    Science.gov (United States)

    Acharya, U Rajendra; Oh, Shu Lih; Hagiwara, Yuki; Tan, Jen Hong; Adam, Muhammad; Gertych, Arkadiusz; Tan, Ru San

    2017-10-01

    The electrocardiogram (ECG) is a standard test used to monitor the activity of the heart. Many cardiac abnormalities will be manifested in the ECG including arrhythmia which is a general term that refers to an abnormal heart rhythm. The basis of arrhythmia diagnosis is the identification of normal versus abnormal individual heart beats, and their correct classification into different diagnoses, based on ECG morphology. Heartbeats can be sub-divided into five categories namely non-ectopic, supraventricular ectopic, ventricular ectopic, fusion, and unknown beats. It is challenging and time-consuming to distinguish these heartbeats on ECG as these signals are typically corrupted by noise. We developed a 9-layer deep convolutional neural network (CNN) to automatically identify 5 different categories of heartbeats in ECG signals. Our experiment was conducted in original and noise attenuated sets of ECG signals derived from a publicly available database. This set was artificially augmented to even out the number of instances the 5 classes of heartbeats and filtered to remove high-frequency noise. The CNN was trained using the augmented data and achieved an accuracy of 94.03% and 93.47% in the diagnostic classification of heartbeats in original and noise free ECGs, respectively. When the CNN was trained with highly imbalanced data (original dataset), the accuracy of the CNN reduced to 89.07%% and 89.3% in noisy and noise-free ECGs. When properly trained, the proposed CNN model can serve as a tool for screening of ECG to quickly identify different types and frequency of arrhythmic heartbeats. Copyright © 2017 Elsevier Ltd. All rights reserved.

  11. Limitations of a convolution method for modeling geometric uncertainties in radiation therapy. I. The effect of shift invariance

    International Nuclear Information System (INIS)

    Craig, Tim; Battista, Jerry; Van Dyk, Jake

    2003-01-01

    Convolution methods have been used to model the effect of geometric uncertainties on dose delivery in radiation therapy. Convolution assumes shift invariance of the dose distribution. Internal inhomogeneities and surface curvature lead to violations of this assumption. The magnitude of the error resulting from violation of shift invariance is not well documented. This issue is addressed by comparing dose distributions calculated using the Convolution method with dose distributions obtained by Direct Simulation. A comparison of conventional Static dose distributions was also made with Direct Simulation. This analysis was performed for phantom geometries and several clinical tumor sites. A modification to the Convolution method to correct for some of the inherent errors is proposed and tested using example phantoms and patients. We refer to this modified method as the Corrected Convolution. The average maximum dose error in the calculated volume (averaged over different beam arrangements in the various phantom examples) was 21% with the Static dose calculation, 9% with Convolution, and reduced to 5% with the Corrected Convolution. The average maximum dose error in the calculated volume (averaged over four clinical examples) was 9% for the Static method, 13% for Convolution, and 3% for Corrected Convolution. While Convolution can provide a superior estimate of the dose delivered when geometric uncertainties are present, the violation of shift invariance can result in substantial errors near the surface of the patient. The proposed Corrected Convolution modification reduces errors near the surface to 3% or less

  12. Digital Tomosynthesis System Geometry Analysis Using Convolution-Based Blur-and-Add (BAA) Model.

    Science.gov (United States)

    Wu, Meng; Yoon, Sungwon; Solomon, Edward G; Star-Lack, Josh; Pelc, Norbert; Fahrig, Rebecca

    2016-01-01

    Digital tomosynthesis is a three-dimensional imaging technique with a lower radiation dose than computed tomography (CT). Due to the missing data in tomosynthesis systems, out-of-plane structures in the depth direction cannot be completely removed by the reconstruction algorithms. In this work, we analyzed the impulse responses of common tomosynthesis systems on a plane-to-plane basis and proposed a fast and accurate convolution-based blur-and-add (BAA) model to simulate the backprojected images. In addition, the analysis formalism describing the impulse response of out-of-plane structures can be generalized to both rotating and parallel gantries. We implemented a ray tracing forward projection and backprojection (ray-based model) algorithm and the convolution-based BAA model to simulate the shift-and-add (backproject) tomosynthesis reconstructions. The convolution-based BAA model with proper geometry distortion correction provides reasonably accurate estimates of the tomosynthesis reconstruction. A numerical comparison indicates that the simulated images using the two models differ by less than 6% in terms of the root-mean-squared error. This convolution-based BAA model can be used in efficient system geometry analysis, reconstruction algorithm design, out-of-plane artifacts suppression, and CT-tomosynthesis registration.

  13. Convolution product construction of interactions in probabilistic physical models

    International Nuclear Information System (INIS)

    Ratsimbarison, H.M.; Raboanary, R.

    2007-01-01

    This paper aims to give a probabilistic construction of interactions which may be relevant for building physical theories such as interacting quantum field theories. We start with the path integral definition of partition function in quantum field theory which recall us the probabilistic nature of this physical theory. From a Gaussian law considered as free theory, an interacting theory is constructed by nontrivial convolution product between the free theory and an interacting term which is also a probability law. The resulting theory, again a probability law, exhibits two proprieties already present in nowadays theories of interactions such as Gauge theory : the interaction term does not depend on the free term, and two different free theories can be implemented with the same interaction.

  14. Adaptive Correlation Model for Visual Tracking Using Keypoints Matching and Deep Convolutional Feature

    Directory of Open Access Journals (Sweden)

    Yuankun Li

    2018-02-01

    Full Text Available Although correlation filter (CF-based visual tracking algorithms have achieved appealing results, there are still some problems to be solved. When the target object goes through long-term occlusions or scale variation, the correlation model used in existing CF-based algorithms will inevitably learn some non-target information or partial-target information. In order to avoid model contamination and enhance the adaptability of model updating, we introduce the keypoints matching strategy and adjust the model learning rate dynamically according to the matching score. Moreover, the proposed approach extracts convolutional features from a deep convolutional neural network (DCNN to accurately estimate the position and scale of the target. Experimental results demonstrate that the proposed tracker has achieved satisfactory performance in a wide range of challenging tracking scenarios.

  15. Optimized parallel convolutions for non-linear fluid models of tokamak ηi turbulence

    International Nuclear Information System (INIS)

    Milovich, J.L.; Tomaschke, G.; Kerbel, G.D.

    1993-01-01

    Non-linear computational fluid models of plasma turbulence based on spectral methods typically spend a large fraction of the total computing time evaluating convolutions. Usually these convolutions arise from an explicit or semi implicit treatment of the convective non-linearities in the problem. Often the principal convective velocity is perpendicular to magnetic field lines allowing a reduction of the convolution to two dimensions in an appropriate geometry, but beyond this, different models vary widely in the particulars of which mode amplitudes are selectively evolved to get the most efficient representation of the turbulence. As the number of modes in the problem, N, increases, the amount of computation required for this part of the evolution algorithm then scales as N 2 /timestep for a direct or analytic method and N ln N/timestep for a pseudospectral method. The constants of proportionality depend on the particulars of mode selection and determine the size problem for which the method will perform equally. For large enough N, the pseudospectral method performance is always superior, though some problems do not require correspondingly high resolution. Further, the Courant condition for numerical stability requires that the timestep size must decrease proportionately as N increases, thus accentuating the need to have fast methods for larger N problems. The authors have developed a package for the Cray system which performs these convolutions for a rather arbitrary mode selection scheme using either method. The package is highly optimized using a combination of macro and microtasking techniques, as well as vectorization and in some cases assembly coded routines. Parts of the package have also been developed and optimized for the CM200 and CM5 system. Performance comparisons with respect to problem size, parallelization, selection schemes and architecture are presented

  16. Maximal monotone model with delay term of convolution

    Directory of Open Access Journals (Sweden)

    Claude-Henri Lamarque

    2005-01-01

    Full Text Available Mechanical models are governed either by partial differential equations with boundary conditions and initial conditions (e.g., in the frame of continuum mechanics or by ordinary differential equations (e.g., after discretization via Galerkin procedure or directly from the model description with the initial conditions. In order to study dynamical behavior of mechanical systems with a finite number of degrees of freedom including nonsmooth terms (e.g., friction, we consider here problems governed by differential inclusions. To describe effects of particular constitutive laws, we add a delay term. In contrast to previous papers, we introduce delay via a Volterra kernel. We provide existence and uniqueness results by using an Euler implicit numerical scheme; then convergence with its order is established. A few numerical examples are given.

  17. Knowledge Based 3d Building Model Recognition Using Convolutional Neural Networks from LIDAR and Aerial Imageries

    Science.gov (United States)

    Alidoost, F.; Arefi, H.

    2016-06-01

    In recent years, with the development of the high resolution data acquisition technologies, many different approaches and algorithms have been presented to extract the accurate and timely updated 3D models of buildings as a key element of city structures for numerous applications in urban mapping. In this paper, a novel and model-based approach is proposed for automatic recognition of buildings' roof models such as flat, gable, hip, and pyramid hip roof models based on deep structures for hierarchical learning of features that are extracted from both LiDAR and aerial ortho-photos. The main steps of this approach include building segmentation, feature extraction and learning, and finally building roof labeling in a supervised pre-trained Convolutional Neural Network (CNN) framework to have an automatic recognition system for various types of buildings over an urban area. In this framework, the height information provides invariant geometric features for convolutional neural network to localize the boundary of each individual roofs. CNN is a kind of feed-forward neural network with the multilayer perceptron concept which consists of a number of convolutional and subsampling layers in an adaptable structure and it is widely used in pattern recognition and object detection application. Since the training dataset is a small library of labeled models for different shapes of roofs, the computation time of learning can be decreased significantly using the pre-trained models. The experimental results highlight the effectiveness of the deep learning approach to detect and extract the pattern of buildings' roofs automatically considering the complementary nature of height and RGB information.

  18. The Gaussian streaming model and convolution Lagrangian effective field theory

    Energy Technology Data Exchange (ETDEWEB)

    Vlah, Zvonimir [Stanford Institute for Theoretical Physics and Department of Physics, Stanford University, Stanford, CA 94306 (United States); Castorina, Emanuele; White, Martin, E-mail: zvlah@stanford.edu, E-mail: ecastorina@berkeley.edu, E-mail: mwhite@berkeley.edu [Department of Physics, University of California, Berkeley, CA 94720 (United States)

    2016-12-01

    We update the ingredients of the Gaussian streaming model (GSM) for the redshift-space clustering of biased tracers using the techniques of Lagrangian perturbation theory, effective field theory (EFT) and a generalized Lagrangian bias expansion. After relating the GSM to the cumulant expansion, we present new results for the real-space correlation function, mean pairwise velocity and pairwise velocity dispersion including counter terms from EFT and bias terms through third order in the linear density, its leading derivatives and its shear up to second order. We discuss the connection to the Gaussian peaks formalism. We compare the ingredients of the GSM to a suite of large N-body simulations, and show the performance of the theory on the low order multipoles of the redshift-space correlation function and power spectrum. We highlight the importance of a general biasing scheme, which we find to be as important as higher-order corrections due to non-linear evolution for the halos we consider on the scales of interest to us.

  19. Supervised Convolutional Sparse Coding

    KAUST Repository

    Affara, Lama Ahmed

    2018-04-08

    Convolutional Sparse Coding (CSC) is a well-established image representation model especially suited for image restoration tasks. In this work, we extend the applicability of this model by proposing a supervised approach to convolutional sparse coding, which aims at learning discriminative dictionaries instead of purely reconstructive ones. We incorporate a supervised regularization term into the traditional unsupervised CSC objective to encourage the final dictionary elements to be discriminative. Experimental results show that using supervised convolutional learning results in two key advantages. First, we learn more semantically relevant filters in the dictionary and second, we achieve improved image reconstruction on unseen data.

  20. Convolution Model of a Queueing System with the cFIFO Service Discipline

    Directory of Open Access Journals (Sweden)

    Sławomir Hanczewski

    2016-01-01

    Full Text Available This article presents an approximate convolution model of a multiservice queueing system with the continuous FIFO (cFIFO service discipline. The model makes it possible to service calls sequentially with variable bit rate, determined by unoccupied (free resources of the multiservice server. As compared to the FIFO discipline, the cFIFO queue utilizes the resources of a multiservice server more effectively. The assumption in the model is that the queueing system is offered a mixture of independent multiservice Bernoulli-Poisson-Pascal (BPP call streams. The article also discusses the results of modelling a number of queueing systems to which different, non-Poissonian, call streams are offered. To verify the accuracy of the model, the results of the analytical calculations are compared with the results of simulation experiments for a number of selected queueing systems. The study has confirmed the accuracy of all adopted theoretical assumptions for the proposed analytical model.

  1. Recurrent Convolutional Neural Networks: A Better Model of Biological Object Recognition.

    Science.gov (United States)

    Spoerer, Courtney J; McClure, Patrick; Kriegeskorte, Nikolaus

    2017-01-01

    Feedforward neural networks provide the dominant model of how the brain performs visual object recognition. However, these networks lack the lateral and feedback connections, and the resulting recurrent neuronal dynamics, of the ventral visual pathway in the human and non-human primate brain. Here we investigate recurrent convolutional neural networks with bottom-up (B), lateral (L), and top-down (T) connections. Combining these types of connections yields four architectures (B, BT, BL, and BLT), which we systematically test and compare. We hypothesized that recurrent dynamics might improve recognition performance in the challenging scenario of partial occlusion. We introduce two novel occluded object recognition tasks to test the efficacy of the models, digit clutter (where multiple target digits occlude one another) and digit debris (where target digits are occluded by digit fragments). We find that recurrent neural networks outperform feedforward control models (approximately matched in parametric complexity) at recognizing objects, both in the absence of occlusion and in all occlusion conditions. Recurrent networks were also found to be more robust to the inclusion of additive Gaussian noise. Recurrent neural networks are better in two respects: (1) they are more neurobiologically realistic than their feedforward counterparts; (2) they are better in terms of their ability to recognize objects, especially under challenging conditions. This work shows that computer vision can benefit from using recurrent convolutional architectures and suggests that the ubiquitous recurrent connections in biological brains are essential for task performance.

  2. Electroencephalography Based Fusion Two-Dimensional (2D-Convolution Neural Networks (CNN Model for Emotion Recognition System

    Directory of Open Access Journals (Sweden)

    Yea-Hoon Kwon

    2018-04-01

    Full Text Available The purpose of this study is to improve human emotional classification accuracy using a convolution neural networks (CNN model and to suggest an overall method to classify emotion based on multimodal data. We improved classification performance by combining electroencephalogram (EEG and galvanic skin response (GSR signals. GSR signals are preprocessed using by the zero-crossing rate. Sufficient EEG feature extraction can be obtained through CNN. Therefore, we propose a suitable CNN model for feature extraction by tuning hyper parameters in convolution filters. The EEG signal is preprocessed prior to convolution by a wavelet transform while considering time and frequency simultaneously. We use a database for emotion analysis using the physiological signals open dataset to verify the proposed process, achieving 73.4% accuracy, showing significant performance improvement over the current best practice models.

  3. Dispersion-convolution model for simulating peaks in a flow injection system.

    Science.gov (United States)

    Pai, Su-Cheng; Lai, Yee-Hwong; Chiao, Ling-Yun; Yu, Tiing

    2007-01-12

    A dispersion-convolution model is proposed for simulating peak shapes in a single-line flow injection system. It is based on the assumption that an injected sample plug is expanded due to a "bulk" dispersion mechanism along the length coordinate, and that after traveling over a distance or a period of time, the sample zone will develop into a Gaussian-like distribution. This spatial pattern is further transformed to a temporal coordinate by a convolution process, and finally a temporal peak image is generated. The feasibility of the proposed model has been examined by experiments with various coil lengths, sample sizes and pumping rates. An empirical dispersion coefficient (D*) can be estimated by using the observed peak position, height and area (tp*, h* and At*) from a recorder. An empirical temporal shift (Phi*) can be further approximated by Phi*=D*/u2, which becomes an important parameter in the restoration of experimental peaks. Also, the dispersion coefficient can be expressed as a second-order polynomial function of the pumping rate Q, for which D*(Q)=delta0+delta1Q+delta2Q2. The optimal dispersion occurs at a pumping rate of Qopt=sqrt[delta0/delta2]. This explains the interesting "Nike-swoosh" relationship between the peak height and pumping rate. The excellent coherence of theoretical and experimental peak shapes confirms that the temporal distortion effect is the dominating reason to explain the peak asymmetry in flow injection analysis.

  4. Glue detection based on teaching points constraint and tracking model of pixel convolution

    Science.gov (United States)

    Geng, Lei; Ma, Xiao; Xiao, Zhitao; Wang, Wen

    2018-01-01

    On-line glue detection based on machine version is significant for rust protection and strengthening in car production. Shadow stripes caused by reflect light and unevenness of inside front cover of car reduce the accuracy of glue detection. In this paper, we propose an effective algorithm to distinguish the edges of the glue and shadow stripes. Teaching points are utilized to calculate slope between the two adjacent points. Then a tracking model based on pixel convolution along motion direction is designed to segment several local rectangular regions using distance. The distance is the height of rectangular region. The pixel convolution along the motion direction is proposed to extract edges of gules in local rectangular region. A dataset with different illumination and complexity shape stripes are used to evaluate proposed method, which include 500 thousand images captured from the camera of glue gun machine. Experimental results demonstrate that the proposed method can detect the edges of glue accurately. The shadow stripes are distinguished and removed effectively. Our method achieves the 99.9% accuracies for the image dataset.

  5. Depth Reconstruction from Single Images Using a Convolutional Neural Network and a Condition Random Field Model.

    Science.gov (United States)

    Liu, Dan; Liu, Xuejun; Wu, Yiguang

    2018-04-24

    This paper presents an effective approach for depth reconstruction from a single image through the incorporation of semantic information and local details from the image. A unified framework for depth acquisition is constructed by joining a deep Convolutional Neural Network (CNN) and a continuous pairwise Conditional Random Field (CRF) model. Semantic information and relative depth trends of local regions inside the image are integrated into the framework. A deep CNN network is firstly used to automatically learn a hierarchical feature representation of the image. To get more local details in the image, the relative depth trends of local regions are incorporated into the network. Combined with semantic information of the image, a continuous pairwise CRF is then established and is used as the loss function of the unified model. Experiments on real scenes demonstrate that the proposed approach is effective and that the approach obtains satisfactory results.

  6. Depth Reconstruction from Single Images Using a Convolutional Neural Network and a Condition Random Field Model

    Directory of Open Access Journals (Sweden)

    Dan Liu

    2018-04-01

    Full Text Available This paper presents an effective approach for depth reconstruction from a single image through the incorporation of semantic information and local details from the image. A unified framework for depth acquisition is constructed by joining a deep Convolutional Neural Network (CNN and a continuous pairwise Conditional Random Field (CRF model. Semantic information and relative depth trends of local regions inside the image are integrated into the framework. A deep CNN network is firstly used to automatically learn a hierarchical feature representation of the image. To get more local details in the image, the relative depth trends of local regions are incorporated into the network. Combined with semantic information of the image, a continuous pairwise CRF is then established and is used as the loss function of the unified model. Experiments on real scenes demonstrate that the proposed approach is effective and that the approach obtains satisfactory results.

  7. A scatter model for fast neutron beams using convolution of diffusion kernels

    International Nuclear Information System (INIS)

    Moyers, M.F.; Horton, J.L.; Boyer, A.L.

    1988-01-01

    A new model is proposed to calculate dose distributions in materials irradiated with fast neutron beams. Scattered neutrons are transported away from the point of production within the irradiated material in the forward, lateral and backward directions, while recoil protons are transported in the forward and lateral directions. The calculation of dose distributions, such as for radiotherapy planning, is accomplished by convolving a primary attenuation distribution with a diffusion kernel. The primary attenuation distribution may be quickly calculated for any given set of beam and material conditions as it describes only the magnitude and distribution of first interaction sites. The calculation of energy diffusion kernels is very time consuming but must be calculated only once for a given energy. Energy diffusion distributions shown in this paper have been calculated using a Monte Carlo type of program. To decrease beam calculation time, convolutions are performed using a Fast Fourier Transform technique. (author)

  8. Strong convective storm nowcasting using a hybrid approach of convolutional neural network and hidden Markov model

    Science.gov (United States)

    Zhang, Wei; Jiang, Ling; Han, Lei

    2018-04-01

    Convective storm nowcasting refers to the prediction of the convective weather initiation, development, and decay in a very short term (typically 0 2 h) .Despite marked progress over the past years, severe convective storm nowcasting still remains a challenge. With the boom of machine learning, it has been well applied in various fields, especially convolutional neural network (CNN). In this paper, we build a servere convective weather nowcasting system based on CNN and hidden Markov model (HMM) using reanalysis meteorological data. The goal of convective storm nowcasting is to predict if there is a convective storm in 30min. In this paper, we compress the VDRAS reanalysis data to low-dimensional data by CNN as the observation vector of HMM, then obtain the development trend of strong convective weather in the form of time series. It shows that, our method can extract robust features without any artificial selection of features, and can capture the development trend of strong convective storm.

  9. Convolution based profile fitting

    International Nuclear Information System (INIS)

    Kern, A.; Coelho, A.A.; Cheary, R.W.

    2002-01-01

    Full text: In convolution based profile fitting, profiles are generated by convoluting functions together to form the observed profile shape. For a convolution of 'n' functions this process can be written as, Y(2θ)=F 1 (2θ)x F 2 (2θ)x... x F i (2θ)x....xF n (2θ). In powder diffractometry the functions F i (2θ) can be interpreted as the aberration functions of the diffractometer, but in general any combination of appropriate functions for F i (2θ) may be used in this context. Most direct convolution fitting methods are restricted to combinations of F i (2θ) that can be convoluted analytically (e.g. GSAS) such as Lorentzians, Gaussians, the hat (impulse) function and the exponential function. However, software such as TOPAS is now available that can accurately convolute and refine a wide variety of profile shapes numerically, including user defined profiles, without the need to convolute analytically. Some of the most important advantages of modern convolution based profile fitting are: 1) virtually any peak shape and angle dependence can normally be described using minimal profile parameters in laboratory and synchrotron X-ray data as well as in CW and TOF neutron data. This is possible because numerical convolution and numerical differentiation is used within the refinement procedure so that a wide range of functions can easily be incorporated into the convolution equation; 2) it can use physically based diffractometer models by convoluting the instrument aberration functions. This can be done for most laboratory based X-ray powder diffractometer configurations including conventional divergent beam instruments, parallel beam instruments, and diffractometers used for asymmetric diffraction. It can also accommodate various optical elements (e.g. multilayers and monochromators) and detector systems (e.g. point and position sensitive detectors) and has already been applied to neutron powder diffraction systems (e.g. ANSTO) as well as synchrotron based

  10. Modeling Habitat Suitability of Migratory Birds from Remote Sensing Images Using Convolutional Neural Networks

    Science.gov (United States)

    Su, Jin-He; Piao, Ying-Chao; Luo, Ze; Yan, Bao-Ping

    2018-01-01

    Simple Summary The understanding of the spatio-temporal distribution of the species habitats would facilitate wildlife resource management and conservation efforts. Existing methods have poor performance due to the limited availability of training samples. More recently, location-aware sensors have been widely used to track animal movements. The aim of the study was to generate suitability maps of bar-head geese using movement data coupled with environmental parameters, such as remote sensing images and temperature data. Therefore, we modified a deep convolutional neural network for the multi-scale inputs. The results indicate that the proposed method can identify the areas with the dense goose species around Qinghai Lake. In addition, this approach might also be interesting for implementation in other species with different niche factors or in areas where biological survey data are scarce. Abstract With the application of various data acquisition devices, a large number of animal movement data can be used to label presence data in remote sensing images and predict species distribution. In this paper, a two-stage classification approach for combining movement data and moderate-resolution remote sensing images was proposed. First, we introduced a new density-based clustering method to identify stopovers from migratory birds’ movement data and generated classification samples based on the clustering result. We split the remote sensing images into 16 × 16 patches and labeled them as positive samples if they have overlap with stopovers. Second, a multi-convolution neural network model is proposed for extracting the features from temperature data and remote sensing images, respectively. Then a Support Vector Machines (SVM) model was used to combine the features together and predict classification results eventually. The experimental analysis was carried out on public Landsat 5 TM images and a GPS dataset was collected on 29 birds over three years. The results

  11. Phase Diagrams of Three-Dimensional Anderson and Quantum Percolation Models Using Deep Three-Dimensional Convolutional Neural Network

    Science.gov (United States)

    Mano, Tomohiro; Ohtsuki, Tomi

    2017-11-01

    The three-dimensional Anderson model is a well-studied model of disordered electron systems that shows the delocalization-localization transition. As in our previous papers on two- and three-dimensional (2D, 3D) quantum phase transitions [J. Phys. Soc. Jpn. 85, 123706 (2016), 86, 044708 (2017)], we used an image recognition algorithm based on a multilayered convolutional neural network. However, in contrast to previous papers in which 2D image recognition was used, we applied 3D image recognition to analyze entire 3D wave functions. We show that a full phase diagram of the disorder-energy plane is obtained once the 3D convolutional neural network has been trained at the band center. We further demonstrate that the full phase diagram for 3D quantum bond and site percolations can be drawn by training the 3D Anderson model at the band center.

  12. Objects Classification by Learning-Based Visual Saliency Model and Convolutional Neural Network.

    Science.gov (United States)

    Li, Na; Zhao, Xinbo; Yang, Yongjia; Zou, Xiaochun

    2016-01-01

    Humans can easily classify different kinds of objects whereas it is quite difficult for computers. As a hot and difficult problem, objects classification has been receiving extensive interests with broad prospects. Inspired by neuroscience, deep learning concept is proposed. Convolutional neural network (CNN) as one of the methods of deep learning can be used to solve classification problem. But most of deep learning methods, including CNN, all ignore the human visual information processing mechanism when a person is classifying objects. Therefore, in this paper, inspiring the completed processing that humans classify different kinds of objects, we bring forth a new classification method which combines visual attention model and CNN. Firstly, we use the visual attention model to simulate the processing of human visual selection mechanism. Secondly, we use CNN to simulate the processing of how humans select features and extract the local features of those selected areas. Finally, not only does our classification method depend on those local features, but also it adds the human semantic features to classify objects. Our classification method has apparently advantages in biology. Experimental results demonstrated that our method made the efficiency of classification improve significantly.

  13. Modeling Habitat Suitability of Migratory Birds from Remote Sensing Images Using Convolutional Neural Networks.

    Science.gov (United States)

    Su, Jin-He; Piao, Ying-Chao; Luo, Ze; Yan, Bao-Ping

    2018-04-26

    With the application of various data acquisition devices, a large number of animal movement data can be used to label presence data in remote sensing images and predict species distribution. In this paper, a two-stage classification approach for combining movement data and moderate-resolution remote sensing images was proposed. First, we introduced a new density-based clustering method to identify stopovers from migratory birds’ movement data and generated classification samples based on the clustering result. We split the remote sensing images into 16 × 16 patches and labeled them as positive samples if they have overlap with stopovers. Second, a multi-convolution neural network model is proposed for extracting the features from temperature data and remote sensing images, respectively. Then a Support Vector Machines (SVM) model was used to combine the features together and predict classification results eventually. The experimental analysis was carried out on public Landsat 5 TM images and a GPS dataset was collected on 29 birds over three years. The results indicated that our proposed method outperforms the existing baseline methods and was able to achieve good performance in habitat suitability prediction.

  14. [Computer aided diagnosis model for lung tumor based on ensemble convolutional neural network].

    Science.gov (United States)

    Wang, Yuanyuan; Zhou, Tao; Lu, Huiling; Wu, Cuiying; Yang, Pengfei

    2017-08-01

    The convolutional neural network (CNN) could be used on computer-aided diagnosis of lung tumor with positron emission tomography (PET)/computed tomography (CT), which can provide accurate quantitative analysis to compensate for visual inertia and defects in gray-scale sensitivity, and help doctors diagnose accurately. Firstly, parameter migration method is used to build three CNNs (CT-CNN, PET-CNN, and PET/CT-CNN) for lung tumor recognition in CT, PET, and PET/CT image, respectively. Then, we aimed at CT-CNN to obtain the appropriate model parameters for CNN training through analysis the influence of model parameters such as epochs, batchsize and image scale on recognition rate and training time. Finally, three single CNNs are used to construct ensemble CNN, and then lung tumor PET/CT recognition was completed through relative majority vote method and the performance between ensemble CNN and single CNN was compared. The experiment results show that the ensemble CNN is better than single CNN on computer-aided diagnosis of lung tumor.

  15. Convolution copula econometrics

    CERN Document Server

    Cherubini, Umberto; Mulinacci, Sabrina

    2016-01-01

    This book presents a novel approach to time series econometrics, which studies the behavior of nonlinear stochastic processes. This approach allows for an arbitrary dependence structure in the increments and provides a generalization with respect to the standard linear independent increments assumption of classical time series models. The book offers a solution to the problem of a general semiparametric approach, which is given by a concept called C-convolution (convolution of dependent variables), and the corresponding theory of convolution-based copulas. Intended for econometrics and statistics scholars with a special interest in time series analysis and copula functions (or other nonparametric approaches), the book is also useful for doctoral students with a basic knowledge of copula functions wanting to learn about the latest research developments in the field.

  16. An Ensemble Deep Convolutional Neural Network Model with Improved D-S Evidence Fusion for Bearing Fault Diagnosis.

    Science.gov (United States)

    Li, Shaobo; Liu, Guokai; Tang, Xianghong; Lu, Jianguang; Hu, Jianjun

    2017-07-28

    Intelligent machine health monitoring and fault diagnosis are becoming increasingly important for modern manufacturing industries. Current fault diagnosis approaches mostly depend on expert-designed features for building prediction models. In this paper, we proposed IDSCNN, a novel bearing fault diagnosis algorithm based on ensemble deep convolutional neural networks and an improved Dempster-Shafer theory based evidence fusion. The convolutional neural networks take the root mean square (RMS) maps from the FFT (Fast Fourier Transformation) features of the vibration signals from two sensors as inputs. The improved D-S evidence theory is implemented via distance matrix from evidences and modified Gini Index. Extensive evaluations of the IDSCNN on the Case Western Reserve Dataset showed that our IDSCNN algorithm can achieve better fault diagnosis performance than existing machine learning methods by fusing complementary or conflicting evidences from different models and sensors and adapting to different load conditions.

  17. The application of convolution-based statistical model on the electrical breakdown time delay distributions in neon

    International Nuclear Information System (INIS)

    Maluckov, Cedomir A.; Karamarkovic, Jugoslav P.; Radovic, Miodrag K.; Pejovic, Momcilo M.

    2004-01-01

    The convolution-based model of the electrical breakdown time delay distribution is applied for statistical analysis of experimental results obtained in neon-filled diode tube at 6.5 mbar. At first, the numerical breakdown time delay density distributions are obtained by stochastic modeling as the sum of two independent random variables, the electrical breakdown statistical time delay with exponential, and discharge formative time with Gaussian distribution. Then, the single characteristic breakdown time delay distribution is obtained as the convolution of these two random variables with previously determined parameters. These distributions show good correspondence with the experimental distributions, obtained on the basis of 1000 successive and independent measurements. The shape of distributions is investigated, and corresponding skewness and kurtosis are plotted, in order to follow the transition from Gaussian to exponential distribution

  18. Limitations of a convolution method for modeling geometric uncertainties in radiation therapy: the radiobiological dose-per-fraction effect

    International Nuclear Information System (INIS)

    Song, William; Battista, Jerry; Van Dyk, Jake

    2004-01-01

    The convolution method can be used to model the effect of random geometric uncertainties into planned dose distributions used in radiation treatment planning. This is effectively done by linearly adding infinitesimally small doses, each with a particular geometric offset, over an assumed infinite number of fractions. However, this process inherently ignores the radiobiological dose-per-fraction effect since only the summed physical dose distribution is generated. The resultant potential error on predicted radiobiological outcome [quantified in this work with tumor control probability (TCP), equivalent uniform dose (EUD), normal tissue complication probability (NTCP), and generalized equivalent uniform dose (gEUD)] has yet to be thoroughly quantified. In this work, the results of a Monte Carlo simulation of geometric displacements are compared to those of the convolution method for random geometric uncertainties of 0, 1, 2, 3, 4, and 5 mm (standard deviation). The α/β CTV ratios of 0.8, 1.5, 3, 5, and 10 Gy are used to represent the range of radiation responses for different tumors, whereas a single α/β OAR ratio of 3 Gy is used to represent all the organs at risk (OAR). The analysis is performed on a four-field prostate treatment plan of 18 MV x rays. The fraction numbers are varied from 1-50, with isoeffective adjustments of the corresponding dose-per-fractions to maintain a constant tumor control, using the linear-quadratic cell survival model. The average differences in TCP and EUD of the target, and in NTCP and gEUD of the OAR calculated from the convolution and Monte Carlo methods reduced asymptotically as the total fraction number increased, with the differences reaching negligible levels beyond the treatment fraction number of ≥20. The convolution method generally overestimates the radiobiological indices, as compared to the Monte Carlo method, for the target volume, and underestimates those for the OAR. These effects are interconnected and attributed

  19. Automatic Fabric Defect Detection with a Multi-Scale Convolutional Denoising Autoencoder Network Model.

    Science.gov (United States)

    Mei, Shuang; Wang, Yudan; Wen, Guojun

    2018-04-02

    Fabric defect detection is a necessary and essential step of quality control in the textile manufacturing industry. Traditional fabric inspections are usually performed by manual visual methods, which are low in efficiency and poor in precision for long-term industrial applications. In this paper, we propose an unsupervised learning-based automated approach to detect and localize fabric defects without any manual intervention. This approach is used to reconstruct image patches with a convolutional denoising autoencoder network at multiple Gaussian pyramid levels and to synthesize detection results from the corresponding resolution channels. The reconstruction residual of each image patch is used as the indicator for direct pixel-wise prediction. By segmenting and synthesizing the reconstruction residual map at each resolution level, the final inspection result can be generated. This newly developed method has several prominent advantages for fabric defect detection. First, it can be trained with only a small amount of defect-free samples. This is especially important for situations in which collecting large amounts of defective samples is difficult and impracticable. Second, owing to the multi-modal integration strategy, it is relatively more robust and accurate compared to general inspection methods (the results at each resolution level can be viewed as a modality). Third, according to our results, it can address multiple types of textile fabrics, from simple to more complex. Experimental results demonstrate that the proposed model is robust and yields good overall performance with high precision and acceptable recall rates.

  20. Automatic Fabric Defect Detection with a Multi-Scale Convolutional Denoising Autoencoder Network Model

    Directory of Open Access Journals (Sweden)

    Shuang Mei

    2018-04-01

    Full Text Available Fabric defect detection is a necessary and essential step of quality control in the textile manufacturing industry. Traditional fabric inspections are usually performed by manual visual methods, which are low in efficiency and poor in precision for long-term industrial applications. In this paper, we propose an unsupervised learning-based automated approach to detect and localize fabric defects without any manual intervention. This approach is used to reconstruct image patches with a convolutional denoising autoencoder network at multiple Gaussian pyramid levels and to synthesize detection results from the corresponding resolution channels. The reconstruction residual of each image patch is used as the indicator for direct pixel-wise prediction. By segmenting and synthesizing the reconstruction residual map at each resolution level, the final inspection result can be generated. This newly developed method has several prominent advantages for fabric defect detection. First, it can be trained with only a small amount of defect-free samples. This is especially important for situations in which collecting large amounts of defective samples is difficult and impracticable. Second, owing to the multi-modal integration strategy, it is relatively more robust and accurate compared to general inspection methods (the results at each resolution level can be viewed as a modality. Third, according to our results, it can address multiple types of textile fabrics, from simple to more complex. Experimental results demonstrate that the proposed model is robust and yields good overall performance with high precision and acceptable recall rates.

  1. Modeling cognitive deficits following neurodegenerative diseases and traumatic brain injuries with deep convolutional neural networks.

    Science.gov (United States)

    Lusch, Bethany; Weholt, Jake; Maia, Pedro D; Kutz, J Nathan

    2018-06-01

    The accurate diagnosis and assessment of neurodegenerative disease and traumatic brain injuries (TBI) remain open challenges. Both cause cognitive and functional deficits due to focal axonal swellings (FAS), but it is difficult to deliver a prognosis due to our limited ability to assess damaged neurons at a cellular level in vivo. We simulate the effects of neurodegenerative disease and TBI using convolutional neural networks (CNNs) as our model of cognition. We utilize biophysically relevant statistical data on FAS to damage the connections in CNNs in a functionally relevant way. We incorporate energy constraints on the brain by pruning the CNNs to be less over-engineered. Qualitatively, we demonstrate that damage leads to human-like mistakes. Our experiments also provide quantitative assessments of how accuracy is affected by various types and levels of damage. The deficit resulting from a fixed amount of damage greatly depends on which connections are randomly injured, providing intuition for why it is difficult to predict impairments. There is a large degree of subjectivity when it comes to interpreting cognitive deficits from complex systems such as the human brain. However, we provide important insight and a quantitative framework for disorders in which FAS are implicated. Copyright © 2018 Elsevier Inc. All rights reserved.

  2. Application of deep learning in determining IR precipitation occurrence: a Convolutional Neural Network model

    Science.gov (United States)

    Wang, C.; Hong, Y.

    2017-12-01

    Infrared (IR) information from Geostationary satellites can be used to retrieve precipitation at pretty high spatiotemporal resolutions. Traditional artificial intelligence (AI) methodologies, such as artificial neural networks (ANN), have been designed to build the relationship between near-surface precipitation and manually derived IR features in products including PERSIANN and PERSIANN-CCS. This study builds an automatic precipitation detection model based on IR data using Convolutional Neural Network (CNN) which is implemented by the newly developed deep learning framework, Caffe. The model judges whether there is rain or no rain at pixel level. Compared with traditional ANN methods, CNN can extract features inside the raw data automatically and thoroughly. In this study, IR data from GOES satellites and precipitation estimates from the next generation QPE (Q2) over the central United States are used as inputs and labels, respectively. The whole datasets during the study period (June to August in 2012) are randomly partitioned to three sub datasets (train, validation and test) to establish the model at the spatial resolution of 0.08°×0.08° and the temporal resolution of 1 hour. The experiments show great improvements of CNN in rain identification compared to the widely used IR-based precipitation product, i.e., PERSIANN-CCS. The overall gain in performance is about 30% for critical success index (CSI), 32% for probability of detection (POD) and 12% for false alarm ratio (FAR). Compared to other recent IR-based precipitation retrieval methods (e.g., PERSIANN-DL developed by University of California Irvine), our model is simpler with less parameters, but achieves equally or even better results. CNN has been applied in computer vision domain successfully, and our results prove the method is suitable for IR precipitation detection. Future studies can expand the application of CNN from precipitation occurrence decision to precipitation amount retrieval.

  3. Fundamentals of convolutional coding

    CERN Document Server

    Johannesson, Rolf

    2015-01-01

    Fundamentals of Convolutional Coding, Second Edition, regarded as a bible of convolutional coding brings you a clear and comprehensive discussion of the basic principles of this field * Two new chapters on low-density parity-check (LDPC) convolutional codes and iterative coding * Viterbi, BCJR, BEAST, list, and sequential decoding of convolutional codes * Distance properties of convolutional codes * Includes a downloadable solutions manual

  4. Central focused convolutional neural networks: Developing a data-driven model for lung nodule segmentation.

    Science.gov (United States)

    Wang, Shuo; Zhou, Mu; Liu, Zaiyi; Liu, Zhenyu; Gu, Dongsheng; Zang, Yali; Dong, Di; Gevaert, Olivier; Tian, Jie

    2017-08-01

    Accurate lung nodule segmentation from computed tomography (CT) images is of great importance for image-driven lung cancer analysis. However, the heterogeneity of lung nodules and the presence of similar visual characteristics between nodules and their surroundings make it difficult for robust nodule segmentation. In this study, we propose a data-driven model, termed the Central Focused Convolutional Neural Networks (CF-CNN), to segment lung nodules from heterogeneous CT images. Our approach combines two key insights: 1) the proposed model captures a diverse set of nodule-sensitive features from both 3-D and 2-D CT images simultaneously; 2) when classifying an image voxel, the effects of its neighbor voxels can vary according to their spatial locations. We describe this phenomenon by proposing a novel central pooling layer retaining much information on voxel patch center, followed by a multi-scale patch learning strategy. Moreover, we design a weighted sampling to facilitate the model training, where training samples are selected according to their degree of segmentation difficulty. The proposed method has been extensively evaluated on the public LIDC dataset including 893 nodules and an independent dataset with 74 nodules from Guangdong General Hospital (GDGH). We showed that CF-CNN achieved superior segmentation performance with average dice scores of 82.15% and 80.02% for the two datasets respectively. Moreover, we compared our results with the inter-radiologists consistency on LIDC dataset, showing a difference in average dice score of only 1.98%. Copyright © 2017. Published by Elsevier B.V.

  5. Fast Convolution Module (Fast Convolution Module)

    National Research Council Canada - National Science Library

    Bierens, L

    1997-01-01

    This report describes the design and realisation of a real-time range azimuth compression module, the so-called 'Fast Convolution Module', based on the fast convolution algorithm developed at TNO-FEL...

  6. Two-dimensional exit dosimetry using a liquid-filled electronic portal imaging device and a convolution model

    International Nuclear Information System (INIS)

    Boellaard, Ronald; Herk, Marcel van; Uiterwaal, Hans; Mijnheer, Ben

    1997-01-01

    Background and purpose: To determine the accuracy of two-dimensional exit dose measurements with an electronic portal imaging device, EPID, using a convolution model for a variety of clinically relevant situations. Materials and methods: Exit doses were derived from portal dose images, obtained with a liquid-filled EPID at distances of 50 cm or more behind the patient, by using a convolution model. The resulting on- and off-axis exit dose values were first compared with ionization chamber exit dose measurements for homogeneous and inhomogeneous phantoms in open and wedged 4,8 and 18 MV photon beams. The accuracy of the EPID exit dose measurements was then determined for a number of anthropomorphic phantoms (lung and larynx) irradiated under clinical conditions and for a few patients treated in an 8 MV beam. The latter results were compared with in vivo exit dose measurements using diodes. Results: The exit dose can be determined from portal images with an accuracy of 1.2% (1 SD) compared with ionization chamber measurements for open beams and homogeneous phantoms at all tested beam qualities. In the presence of wedges and for inhomogeneous phantoms the average relative accuracy slightly deteriorated to 1.7% (1 SD). For lung phantoms in a 4 MV beam a similar accuracy was obtained after refinement of our convolution model, which requires knowledge of the patient contour. Differences between diode and EPID exit dose measurements for an anthropomorphic lung phantom in an 8 MV beam were 2.5% at most, with an average agreement within 1% (1 SD). For larynx phantoms in a 4 MV beam exit doses obtained with an ionization chamber and EPID agreed within 1.5% (1 SD). Finally, exit doses in a few patients irradiated in an 8 MV beam could be determined with the EPID with an accuracy of 1.1% (1 SD) relative to exit dose measurements using diodes. Conclusions: Portal images, obtained with our EPID and analyzed with our convolution model, can be used to determine the exit dose

  7. Study of asymmetry in motor areas related to handedness using the fMRI BOLD response Gaussian convolution model

    International Nuclear Information System (INIS)

    Gao Qing; Chen Huafu; Gong Qiyong

    2009-01-01

    Brain asymmetry is a phenomenon well known for handedness, and has been studied in the motor cortex. However, few studies have quantitatively assessed the asymmetrical cortical activities for handedness in motor areas. In the present study, we systematically and quantitatively investigated asymmetry in the left and right primary motor cortices during sequential finger movements using the Gaussian convolution model approach based on the functional magnetic resonance imaging (fMRI) blood oxygenation level dependent (BOLD) response. Six right-handed and six left-handed subjects were recruited to perform three types of hand movement tasks. The results for the expected value of the Gaussian convolution model showed that it took the dominant hand a longer average interval of response delay regardless of the handedness and bi- or uni-manual performance. The results for the standard deviation of the Gaussian model suggested that in the mass neurons, these intervals of the dominant hand were much more variable than those of the non-dominant hand. When comparing bi-manual movement conditions with uni-manual movement conditions in the primary motor cortex (PMC), both the expected value and standard deviation in the Gaussian function were significantly smaller (p < 0.05) in the bi-manual conditions, showing that the movement of the non-dominant hand influenced that of the dominant hand.

  8. Study of asymmetry in motor areas related to handedness using the fMRI BOLD response Gaussian convolution model

    Energy Technology Data Exchange (ETDEWEB)

    Gao Qing [School of Life Science and Technology, University of Electronic Science and Technology of China, Chengdu 610054 (China); School of Applied Mathematics, University of Electronic Science and Technology of China, Chengdu 610054 (China); Chen Huafu [School of Life Science and Technology, University of Electronic Science and Technology of China, Chengdu 610054 (China); School of Applied Mathematics, University of Electronic Science and Technology of China, Chengdu 610054 (China)], E-mail: Chenhf@uestc.edu.cn; Gong Qiyong [Huaxi MR Research Center (HMRRC), Department of Radiology, West China Hospital of Sichuan University, Chengdu 610041 (China)

    2009-10-30

    Brain asymmetry is a phenomenon well known for handedness, and has been studied in the motor cortex. However, few studies have quantitatively assessed the asymmetrical cortical activities for handedness in motor areas. In the present study, we systematically and quantitatively investigated asymmetry in the left and right primary motor cortices during sequential finger movements using the Gaussian convolution model approach based on the functional magnetic resonance imaging (fMRI) blood oxygenation level dependent (BOLD) response. Six right-handed and six left-handed subjects were recruited to perform three types of hand movement tasks. The results for the expected value of the Gaussian convolution model showed that it took the dominant hand a longer average interval of response delay regardless of the handedness and bi- or uni-manual performance. The results for the standard deviation of the Gaussian model suggested that in the mass neurons, these intervals of the dominant hand were much more variable than those of the non-dominant hand. When comparing bi-manual movement conditions with uni-manual movement conditions in the primary motor cortex (PMC), both the expected value and standard deviation in the Gaussian function were significantly smaller (p < 0.05) in the bi-manual conditions, showing that the movement of the non-dominant hand influenced that of the dominant hand.

  9. A convolutional approach to reflection symmetry

    DEFF Research Database (Denmark)

    Cicconet, Marcelo; Birodkar, Vighnesh; Lund, Mads

    2017-01-01

    We present a convolutional approach to reflection symmetry detection in 2D. Our model, built on the products of complex-valued wavelet convolutions, simplifies previous edge-based pairwise methods. Being parameter-centered, as opposed to feature-centered, it has certain computational advantages w...

  10. Exploring the effects of transducer models when training convolutional neural networks to eliminate reflection artifacts in experimental photoacoustic images

    Science.gov (United States)

    Allman, Derek; Reiter, Austin; Bell, Muyinatu

    2018-02-01

    We previously proposed a method of removing reflection artifacts in photoacoustic images that uses deep learning. Our approach generally relies on using simulated photoacoustic channel data to train a convolutional neural network (CNN) that is capable of distinguishing sources from artifacts based on unique differences in their spatial impulse responses (manifested as depth-based differences in wavefront shapes). In this paper, we directly compare a CNN trained with our previous continuous transducer model to a CNN trained with an updated discrete acoustic receiver model that more closely matches an experimental ultrasound transducer. These two CNNs were trained with simulated data and tested on experimental data. The CNN trained using the continuous receiver model correctly classified 100% of sources and 70.3% of artifacts in the experimental data. In contrast, the CNN trained using the discrete receiver model correctly classified 100% of sources and 89.7% of artifacts in the experimental images. The 19.4% increase in artifact classification accuracy indicates that an acoustic receiver model that closely mimics the experimental transducer plays an important role in improving the classification of artifacts in experimental photoacoustic data. Results are promising for developing a method to display CNN-based images that remove artifacts in addition to only displaying network-identified sources as previously proposed.

  11. Application of Convolution Perfectly Matched Layer in MRTD scattering model for non-spherical aerosol particles and its performance analysis

    Science.gov (United States)

    Hu, Shuai; Gao, Taichang; Li, Hao; Yang, Bo; Jiang, Zidong; Liu, Lei; Chen, Ming

    2017-10-01

    The performance of absorbing boundary condition (ABC) is an important factor influencing the simulation accuracy of MRTD (Multi-Resolution Time-Domain) scattering model for non-spherical aerosol particles. To this end, the Convolution Perfectly Matched Layer (CPML), an excellent ABC in FDTD scheme, is generalized and applied to the MRTD scattering model developed by our team. In this model, the time domain is discretized by exponential differential scheme, and the discretization of space domain is implemented by Galerkin principle. To evaluate the performance of CPML, its simulation results are compared with those of BPML (Berenger's Perfectly Matched Layer) and ADE-PML (Perfectly Matched Layer with Auxiliary Differential Equation) for spherical and non-spherical particles, and their simulation errors are analyzed as well. The simulation results show that, for scattering phase matrices, the performance of CPML is better than that of BPML; the computational accuracy of CPML is comparable to that of ADE-PML on the whole, but at scattering angles where phase matrix elements fluctuate sharply, the performance of CPML is slightly better than that of ADE-PML. After orientation averaging process, the differences among the results of different ABCs are reduced to some extent. It also can be found that ABCs have a much weaker influence on integral scattering parameters (such as extinction and absorption efficiencies) than scattering phase matrices, this phenomenon can be explained by the error averaging process in the numerical volume integration.

  12. Prostate segmentation in MRI using a convolutional neural network architecture and training strategy based on statistical shape models.

    Science.gov (United States)

    Karimi, Davood; Samei, Golnoosh; Kesch, Claudia; Nir, Guy; Salcudean, Septimiu E

    2018-05-15

    Most of the existing convolutional neural network (CNN)-based medical image segmentation methods are based on methods that have originally been developed for segmentation of natural images. Therefore, they largely ignore the differences between the two domains, such as the smaller degree of variability in the shape and appearance of the target volume and the smaller amounts of training data in medical applications. We propose a CNN-based method for prostate segmentation in MRI that employs statistical shape models to address these issues. Our CNN predicts the location of the prostate center and the parameters of the shape model, which determine the position of prostate surface keypoints. To train such a large model for segmentation of 3D images using small data (1) we adopt a stage-wise training strategy by first training the network to predict the prostate center and subsequently adding modules for predicting the parameters of the shape model and prostate rotation, (2) we propose a data augmentation method whereby the training images and their prostate surface keypoints are deformed according to the displacements computed based on the shape model, and (3) we employ various regularization techniques. Our proposed method achieves a Dice score of 0.88, which is obtained by using both elastic-net and spectral dropout for regularization. Compared with a standard CNN-based method, our method shows significantly better segmentation performance on the prostate base and apex. Our experiments also show that data augmentation using the shape model significantly improves the segmentation results. Prior knowledge about the shape of the target organ can improve the performance of CNN-based segmentation methods, especially where image features are not sufficient for a precise segmentation. Statistical shape models can also be employed to synthesize additional training data that can ease the training of large CNNs.

  13. Deep Convolutional Neural Networks Outperform Feature-Based But Not Categorical Models in Explaining Object Similarity Judgments

    Science.gov (United States)

    Jozwik, Kamila M.; Kriegeskorte, Nikolaus; Storrs, Katherine R.; Mur, Marieke

    2017-01-01

    Recent advances in Deep convolutional Neural Networks (DNNs) have enabled unprecedentedly accurate computational models of brain representations, and present an exciting opportunity to model diverse cognitive functions. State-of-the-art DNNs achieve human-level performance on object categorisation, but it is unclear how well they capture human behavior on complex cognitive tasks. Recent reports suggest that DNNs can explain significant variance in one such task, judging object similarity. Here, we extend these findings by replicating them for a rich set of object images, comparing performance across layers within two DNNs of different depths, and examining how the DNNs’ performance compares to that of non-computational “conceptual” models. Human observers performed similarity judgments for a set of 92 images of real-world objects. Representations of the same images were obtained in each of the layers of two DNNs of different depths (8-layer AlexNet and 16-layer VGG-16). To create conceptual models, other human observers generated visual-feature labels (e.g., “eye”) and category labels (e.g., “animal”) for the same image set. Feature labels were divided into parts, colors, textures and contours, while category labels were divided into subordinate, basic, and superordinate categories. We fitted models derived from the features, categories, and from each layer of each DNN to the similarity judgments, using representational similarity analysis to evaluate model performance. In both DNNs, similarity within the last layer explains most of the explainable variance in human similarity judgments. The last layer outperforms almost all feature-based models. Late and mid-level layers outperform some but not all feature-based models. Importantly, categorical models predict similarity judgments significantly better than any DNN layer. Our results provide further evidence for commonalities between DNNs and brain representations. Models derived from visual features

  14. Accelerated Time-Domain Modeling of Electromagnetic Pulse Excitation of Finite-Length Dissipative Conductors over a Ground Plane via Function Fitting and Recursive Convolution

    Energy Technology Data Exchange (ETDEWEB)

    Campione, Salvatore [Sandia National Lab. (SNL-NM), Albuquerque, NM (United States); Warne, Larry K. [Sandia National Lab. (SNL-NM), Albuquerque, NM (United States); Sainath, Kamalesh [Sandia National Lab. (SNL-NM), Albuquerque, NM (United States); Basilio, Lorena I. [Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)

    2017-10-01

    In this report we overview the fundamental concepts for a pair of techniques which together greatly hasten computational predictions of electromagnetic pulse (EMP) excitation of finite-length dissipative conductors over a ground plane. In a time- domain, transmission line (TL) model implementation, predictions are computationally bottlenecked time-wise, either for late-time predictions (about 100ns-10000ns range) or predictions concerning EMP excitation of long TLs (order of kilometers or more ). This is because the method requires a temporal convolution to account for the losses in the ground. Addressing this to facilitate practical simulation of EMP excitation of TLs, we first apply a technique to extract an (approximate) complex exponential function basis-fit to the ground/Earth's impedance function, followed by incorporating this into a recursion-based convolution acceleration technique. Because the recursion-based method only requires the evaluation of the most recent voltage history data (versus the entire history in a "brute-force" convolution evaluation), we achieve necessary time speed- ups across a variety of TL/Earth geometry/material scenarios. Intentionally Left Blank

  15. Airplane detection based on fusion framework by combining saliency model with Deep Convolutional Neural Networks

    Science.gov (United States)

    Dou, Hao; Sun, Xiao; Li, Bin; Deng, Qianqian; Yang, Xubo; Liu, Di; Tian, Jinwen

    2018-03-01

    Aircraft detection from very high resolution remote sensing images, has gained more increasing interest in recent years due to the successful civil and military applications. However, several problems still exist: 1) how to extract the high-level features of aircraft; 2) locating objects within such a large image is difficult and time consuming; 3) A common problem of multiple resolutions of satellite images still exists. In this paper, inspirited by biological visual mechanism, the fusion detection framework is proposed, which fusing the top-down visual mechanism (deep CNN model) and bottom-up visual mechanism (GBVS) to detect aircraft. Besides, we use multi-scale training method for deep CNN model to solve the problem of multiple resolutions. Experimental results demonstrate that our method can achieve a better detection result than the other methods.

  16. Convolutive ICA for Spatio-Temporal Analysis of EEG

    DEFF Research Database (Denmark)

    Dyrholm, Mads; Makeig, Scott; Hansen, Lars Kai

    2007-01-01

    in the convolutive model can be correctly detected using Bayesian model selection. We demonstrate a framework for deconvolving an EEG ICA subspace. Initial results suggest that in some cases convolutive mixing may be a more realistic model for EEG signals than the instantaneous ICA model....

  17. Hyper-chaos encryption using convolutional masking and model free unmasking

    International Nuclear Information System (INIS)

    Qi Guo-Yuan; Matondo Sandra Bazebo

    2014-01-01

    In this paper, during the masking process the encrypted message is convolved and embedded into a Qi hyper-chaotic system characterizing a high disorder degree. The masking scheme was tested using both Qi hyper-chaos and Lorenz chaos and indicated that Qi hyper-chaos based masking can resist attacks of the filtering and power spectrum analysis, while the Lorenz based scheme fails for high amplitude data. To unmask the message at the receiving end, two methods are proposed. In the first method, a model-free synchronizer, i.e. a multivariable higher-order differential feedback controller between the transmitter and receiver is employed to de-convolve the message embedded in the receiving signal. In the second method, no synchronization is required since the message is de-convolved using the information of the estimated derivative. (general)

  18. Influence of convolution filtering on coronary plaque attenuation values: observations in an ex vivo model of multislice computed tomography coronary angiography

    International Nuclear Information System (INIS)

    Cademartiri, Filippo; Palumbo, Alessandro; La Grutta, Ludovico; Runza, Giuseppe; Maffei, Erica; Mollet, Nico R.; Hamers, Ronald; Bruining, Nico; Bartolotta, Tommaso V.; Midiri, Massimo; Somers, Pamela; Knaapen, Michiel; Verheye, Stefan

    2007-01-01

    Attenuation variability (measured in Hounsfield Units, HU) of human coronary plaques using multislice computed tomography (MSCT) was evaluated in an ex vivo model with increasing convolution kernels. MSCT was performed in seven ex vivo left coronary arteries sunk into oil followingthe instillation of saline (1/∞) and a 1/50 solution of contrast material (400 mgI/ml iomeprol). Scan parameters were: slices/collimation, 16/0.75 mm; rotation time, 375 ms. Four convolution kernels were used: b30f-smooth, b36f-medium smooth, b46f-medium and b60f-sharp. An experienced radiologist scored for the presence of plaques and measured the attenuation in lumen, calcified and noncalcified plaques and the surrounding oil. The results were compared by the ANOVA test and correlated with Pearson's test. The signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were calculated. The mean attenuation values were significantly different between the four filters (p < 0.0001) in each structure with both solutions. After clustering for the filter, all of the noncalcified plaque values (20.8 ± 39.1, 14.2 ± 35.8, 14.0 ± 32.0, 3.2 ± 32.4 HU with saline; 74.7 ± 66.6, 68.2 ± 63.3, 66.3 ± 66.5, 48.5 ± 60.0 HU in contrast solution) were significantly different, with the exception of the pair b36f-b46f, for which a moderate-high correlation was generally found. Improved SNRs and CNRs were achieved by b30f and b46f. The use of different convolution filters significantly modified the attenuation values, while sharper filtering increased the calcified plaque attenuation and reduced the noncalcified plaque attenuation. (orig.)

  19. Modeling Human Body Using Four-Pole Debye Model in Piecewise Linear Recursive Convolution FDTD Method for the SAR Calculation in the Case of Vehicular Antenna

    Directory of Open Access Journals (Sweden)

    Ammar Guellab

    2018-01-01

    Full Text Available We propose an efficient finite difference time domain (FDTD method based on the piecewise linear recursive convolution (PLRC technique to evaluate the human body exposure to electromagnetic (EM radiation. The source of radiation considered in this study is a high-power antenna, mounted on a military vehicle, covering a broad band of frequency (100 MHz–3 GHz. The simulation is carried out using a nonhomogeneous human body model which takes into consideration most of the internal body tissues. The human tissues are modeled by a four-pole Debye model which is derived from experimental data by using particle swarm optimization (PSO. The human exposure to EM radiation is evaluated by computing the local and whole-body average specific absorption rate (SAR for each occupant. The higher in-tissue electric field intensity points are localized, and the SAR values are compared with the crew safety standard recommendations. The accuracy of the proposed PLRC-FDTD approach and the matching of the Debye model with the experimental data are verified in this study.

  20. Efficient convolutional sparse coding

    Science.gov (United States)

    Wohlberg, Brendt

    2017-06-20

    Computationally efficient algorithms may be applied for fast dictionary learning solving the convolutional sparse coding problem in the Fourier domain. More specifically, efficient convolutional sparse coding may be derived within an alternating direction method of multipliers (ADMM) framework that utilizes fast Fourier transforms (FFT) to solve the main linear system in the frequency domain. Such algorithms may enable a significant reduction in computational cost over conventional approaches by implementing a linear solver for the most critical and computationally expensive component of the conventional iterative algorithm. The theoretical computational cost of the algorithm may be reduced from O(M.sup.3N) to O(MN log N), where N is the dimensionality of the data and M is the number of elements in the dictionary. This significant improvement in efficiency may greatly increase the range of problems that can practically be addressed via convolutional sparse representations.

  1. Multithreaded implicitly dealiased convolutions

    Science.gov (United States)

    Roberts, Malcolm; Bowman, John C.

    2018-03-01

    Implicit dealiasing is a method for computing in-place linear convolutions via fast Fourier transforms that decouples work memory from input data. It offers easier memory management and, for long one-dimensional input sequences, greater efficiency than conventional zero-padding. Furthermore, for convolutions of multidimensional data, the segregation of data and work buffers can be exploited to reduce memory usage and execution time significantly. This is accomplished by processing and discarding data as it is generated, allowing work memory to be reused, for greater data locality and performance. A multithreaded implementation of implicit dealiasing that accepts an arbitrary number of input and output vectors and a general multiplication operator is presented, along with an improved one-dimensional Hermitian convolution that avoids the loop dependency inherent in previous work. An alternate data format that can accommodate a Nyquist mode and enhance cache efficiency is also proposed.

  2. TernaryNet: faster deep model inference without GPUs for medical 3D segmentation using sparse and binary convolutions.

    Science.gov (United States)

    Heinrich, Mattias P; Blendowski, Max; Oktay, Ozan

    2018-05-30

    Deep convolutional neural networks (DCNN) are currently ubiquitous in medical imaging. While their versatility and high-quality results for common image analysis tasks including segmentation, localisation and prediction is astonishing, the large representational power comes at the cost of highly demanding computational effort. This limits their practical applications for image-guided interventions and diagnostic (point-of-care) support using mobile devices without graphics processing units (GPU). We propose a new scheme that approximates both trainable weights and neural activations in deep networks by ternary values and tackles the open question of backpropagation when dealing with non-differentiable functions. Our solution enables the removal of the expensive floating-point matrix multiplications throughout any convolutional neural network and replaces them by energy- and time-preserving binary operators and population counts. We evaluate our approach for the segmentation of the pancreas in CT. Here, our ternary approximation within a fully convolutional network leads to more than 90% memory reductions and high accuracy (without any post-processing) with a Dice overlap of 71.0% that comes close to the one obtained when using networks with high-precision weights and activations. We further provide a concept for sub-second inference without GPUs and demonstrate significant improvements in comparison with binary quantisation and without our proposed ternary hyperbolic tangent continuation. We present a key enabling technique for highly efficient DCNN inference without GPUs that will help to bring the advances of deep learning to practical clinical applications. It has also great promise for improving accuracies in large-scale medical data retrieval.

  3. Nuclear norm regularized convolutional Max Pos@Top machine

    KAUST Repository

    Li, Qinfeng; Zhou, Xiaofeng; Gu, Aihua; Li, Zonghua; Liang, Ru-Ze

    2016-01-01

    , named as Pos@Top. Our proposed classification model has a convolutional structure that is composed by four layers, i.e., the convolutional layer, the activation layer, the max-pooling layer and the full connection layer. In this paper, we propose

  4. Anti-3D Weapon Model Detection for Safe 3D Printing Based on Convolutional Neural Networks and D2 Shape Distribution

    Directory of Open Access Journals (Sweden)

    Giao N. Pham

    2018-03-01

    Full Text Available With the development of 3D printing, weapons are easily printed without any restriction from the production managers. Therefore, anti-3D weapon model detection is necessary issue in safe 3D printing to prevent the printing of 3D weapon models. In this paper, we would like to propose an anti-3D weapon model detection algorithm to prevent the printing of anti-3D weapon models for safe 3D printing based on the D2 shape distribution and an improved convolutional neural networks (CNNs. The purpose of the proposed algorithm is to detect anti-3D weapon models when they are used in 3D printing. The D2 shape distribution is computed from random points on the surface of a 3D weapon model and their geometric features in order to construct a D2 vector. The D2 vector is then trained by improved CNNs. The CNNs are used to detect anti-3D weapon models for safe 3D printing by training D2 vectors which have been constructed from the D2 shape distribution of 3D weapon models. Experiments with 3D weapon models proved that the D2 shape distribution of 3D weapon models in the same class is the same. Training and testing results also verified that the accuracy of the proposed algorithm is higher than the conventional works. The proposed algorithm is applied in a small application, and it could detect anti-3D weapon models for safe 3D printing.

  5. Convolutional Neural Networks - Generalizability and Interpretations

    DEFF Research Database (Denmark)

    Malmgren-Hansen, David

    from data despite it being limited in amount or context representation. Within Machine Learning this thesis focuses on Convolutional Neural Networks for Computer Vision. The research aims to answer how to explore a model's generalizability to the whole population of data samples and how to interpret...

  6. Convolution-deconvolution in DIGES

    International Nuclear Information System (INIS)

    Philippacopoulos, A.J.; Simos, N.

    1995-01-01

    Convolution and deconvolution operations is by all means a very important aspect of SSI analysis since it influences the input to the seismic analysis. This paper documents some of the convolution/deconvolution procedures which have been implemented into the DIGES code. The 1-D propagation of shear and dilatational waves in typical layered configurations involving a stack of layers overlying a rock is treated by DIGES in a similar fashion to that of available codes, e.g. CARES, SHAKE. For certain configurations, however, there is no need to perform such analyses since the corresponding solutions can be obtained in analytic form. Typical cases involve deposits which can be modeled by a uniform halfspace or simple layered halfspaces. For such cases DIGES uses closed-form solutions. These solutions are given for one as well as two dimensional deconvolution. The type of waves considered include P, SV and SH waves. The non-vertical incidence is given special attention since deconvolution can be defined differently depending on the problem of interest. For all wave cases considered, corresponding transfer functions are presented in closed-form. Transient solutions are obtained in the frequency domain. Finally, a variety of forms are considered for representing the free field motion both in terms of deterministic as well as probabilistic representations. These include (a) acceleration time histories, (b) response spectra (c) Fourier spectra and (d) cross-spectral densities

  7. Convolutional coding techniques for data protection

    Science.gov (United States)

    Massey, J. L.

    1975-01-01

    Results of research on the use of convolutional codes in data communications are presented. Convolutional coding fundamentals are discussed along with modulation and coding interaction. Concatenated coding systems and data compression with convolutional codes are described.

  8. The convolution transform

    CERN Document Server

    Hirschman, Isidore Isaac

    2005-01-01

    In studies of general operators of the same nature, general convolution transforms are immediately encountered as the objects of inversion. The relation between differential operators and integral transforms is the basic theme of this work, which is geared toward upper-level undergraduates and graduate students. It may be read easily by anyone with a working knowledge of real and complex variable theory. Topics include the finite and non-finite kernels, variation diminishing transforms, asymptotic behavior of kernels, real inversion theory, representation theory, the Weierstrass transform, and

  9. Towards dropout training for convolutional neural networks.

    Science.gov (United States)

    Wu, Haibing; Gu, Xiaodong

    2015-11-01

    Recently, dropout has seen increasing use in deep learning. For deep convolutional neural networks, dropout is known to work well in fully-connected layers. However, its effect in convolutional and pooling layers is still not clear. This paper demonstrates that max-pooling dropout is equivalent to randomly picking activation based on a multinomial distribution at training time. In light of this insight, we advocate employing our proposed probabilistic weighted pooling, instead of commonly used max-pooling, to act as model averaging at test time. Empirical evidence validates the superiority of probabilistic weighted pooling. We also empirically show that the effect of convolutional dropout is not trivial, despite the dramatically reduced possibility of over-fitting due to the convolutional architecture. Elaborately designing dropout training simultaneously in max-pooling and fully-connected layers, we achieve state-of-the-art performance on MNIST, and very competitive results on CIFAR-10 and CIFAR-100, relative to other approaches without data augmentation. Finally, we compare max-pooling dropout and stochastic pooling, both of which introduce stochasticity based on multinomial distributions at pooling stage. Copyright © 2015 Elsevier Ltd. All rights reserved.

  10. Separating Underdetermined Convolutive Speech Mixtures

    DEFF Research Database (Denmark)

    Pedersen, Michael Syskind; Wang, DeLiang; Larsen, Jan

    2006-01-01

    a method for underdetermined blind source separation of convolutive mixtures. The proposed framework is applicable for separation of instantaneous as well as convolutive speech mixtures. It is possible to iteratively extract each speech signal from the mixture by combining blind source separation...

  11. Strongly-MDS convolutional codes

    NARCIS (Netherlands)

    Gluesing-Luerssen, H; Rosenthal, J; Smarandache, R

    Maximum-distance separable (MDS) convolutional codes have the property that their free distance is maximal among all codes of the same rate and the same degree. In this paper, a class of MDS convolutional codes is introduced whose column distances reach the generalized Singleton bound at the

  12. Consensus Convolutional Sparse Coding

    KAUST Repository

    Choudhury, Biswarup

    2017-12-01

    Convolutional sparse coding (CSC) is a promising direction for unsupervised learning in computer vision. In contrast to recent supervised methods, CSC allows for convolutional image representations to be learned that are equally useful for high-level vision tasks and low-level image reconstruction and can be applied to a wide range of tasks without problem-specific retraining. Due to their extreme memory requirements, however, existing CSC solvers have so far been limited to low-dimensional problems and datasets using a handful of low-resolution example images at a time. In this paper, we propose a new approach to solving CSC as a consensus optimization problem, which lifts these limitations. By learning CSC features from large-scale image datasets for the first time, we achieve significant quality improvements in a number of imaging tasks. Moreover, the proposed method enables new applications in high-dimensional feature learning that has been intractable using existing CSC methods. This is demonstrated for a variety of reconstruction problems across diverse problem domains, including 3D multispectral demosaicing and 4D light field view synthesis.

  13. Consensus Convolutional Sparse Coding

    KAUST Repository

    Choudhury, Biswarup

    2017-04-11

    Convolutional sparse coding (CSC) is a promising direction for unsupervised learning in computer vision. In contrast to recent supervised methods, CSC allows for convolutional image representations to be learned that are equally useful for high-level vision tasks and low-level image reconstruction and can be applied to a wide range of tasks without problem-specific retraining. Due to their extreme memory requirements, however, existing CSC solvers have so far been limited to low-dimensional problems and datasets using a handful of low-resolution example images at a time. In this paper, we propose a new approach to solving CSC as a consensus optimization problem, which lifts these limitations. By learning CSC features from large-scale image datasets for the first time, we achieve significant quality improvements in a number of imaging tasks. Moreover, the proposed method enables new applications in high dimensional feature learning that has been intractable using existing CSC methods. This is demonstrated for a variety of reconstruction problems across diverse problem domains, including 3D multispectral demosaickingand 4D light field view synthesis.

  14. Consensus Convolutional Sparse Coding

    KAUST Repository

    Choudhury, Biswarup; Swanson, Robin; Heide, Felix; Wetzstein, Gordon; Heidrich, Wolfgang

    2017-01-01

    Convolutional sparse coding (CSC) is a promising direction for unsupervised learning in computer vision. In contrast to recent supervised methods, CSC allows for convolutional image representations to be learned that are equally useful for high-level vision tasks and low-level image reconstruction and can be applied to a wide range of tasks without problem-specific retraining. Due to their extreme memory requirements, however, existing CSC solvers have so far been limited to low-dimensional problems and datasets using a handful of low-resolution example images at a time. In this paper, we propose a new approach to solving CSC as a consensus optimization problem, which lifts these limitations. By learning CSC features from large-scale image datasets for the first time, we achieve significant quality improvements in a number of imaging tasks. Moreover, the proposed method enables new applications in high-dimensional feature learning that has been intractable using existing CSC methods. This is demonstrated for a variety of reconstruction problems across diverse problem domains, including 3D multispectral demosaicing and 4D light field view synthesis.

  15. Deep Recurrent Convolutional Neural Network: Improving Performance For Speech Recognition

    OpenAIRE

    Zhang, Zewang; Sun, Zheng; Liu, Jiaqi; Chen, Jingwen; Huo, Zhao; Zhang, Xiao

    2016-01-01

    A deep learning approach has been widely applied in sequence modeling problems. In terms of automatic speech recognition (ASR), its performance has significantly been improved by increasing large speech corpus and deeper neural network. Especially, recurrent neural network and deep convolutional neural network have been applied in ASR successfully. Given the arising problem of training speed, we build a novel deep recurrent convolutional network for acoustic modeling and then apply deep resid...

  16. Dealiased convolutions for pseudospectral simulations

    International Nuclear Information System (INIS)

    Roberts, Malcolm; Bowman, John C

    2011-01-01

    Efficient algorithms have recently been developed for calculating dealiased linear convolution sums without the expense of conventional zero-padding or phase-shift techniques. For one-dimensional in-place convolutions, the memory requirements are identical with the zero-padding technique, with the important distinction that the additional work memory need not be contiguous with the input data. This decoupling of data and work arrays dramatically reduces the memory and computation time required to evaluate higher-dimensional in-place convolutions. The memory savings is achieved by computing the in-place Fourier transform of the data in blocks, rather than all at once. The technique also allows one to dealias the n-ary convolutions that arise on Fourier transforming cubic and higher powers. Implicitly dealiased convolutions can be built on top of state-of-the-art adaptive fast Fourier transform libraries like FFTW. Vectorized multidimensional implementations for the complex and centered Hermitian (pseudospectral) cases have already been implemented in the open-source software FFTW++. With the advent of this library, writing a high-performance dealiased pseudospectral code for solving nonlinear partial differential equations has now become a relatively straightforward exercise. New theoretical estimates of computational complexity and memory use are provided, including corrected timing results for 3D pruned convolutions and further consideration of higher-order convolutions.

  17. Prediction of Electricity Usage Using Convolutional Neural Networks

    OpenAIRE

    Hansen, Martin

    2017-01-01

    Master's thesis Information- and communication technology IKT590 - University of Agder 2017 Convolutional Neural Networks are overwhelmingly accurate when attempting to predict numbers using the famous MNIST-dataset. In this paper, we are attempting to transcend these results for time- series forecasting, and compare them with several regression mod- els. The Convolutional Neural Network model predicted the same value through the entire time lapse in contrast with the other ...

  18. Influence of increasing convolution kernel filtering on plaque imaging with multislice CT using an ex-vivo model of Coronary Angiography

    International Nuclear Information System (INIS)

    Cademartiri, Filippo; Mollet, Nico R.; Runza, Giuseppe

    2005-01-01

    Purpose. To assess the variability in attenuation of coronary plaques with multislice CT angiography (MSCT-CA) in an ex-vivo model with varying convolution kernels. Materials and methods. MSCT-CA (Sensation 16, Siemens) was performed in three ex-vivo left coronary arteries after instillation of contrast material solution (Iomeprol 400 mgI/ml, dilution: 1180). The specimens were placed in oil to simulate epicardial fat. Scan parameters: slices 16/0.75 mm, rotation time 375 ms, feed/rotation 3.0 mm, mAs 500, slice thickness 1 mm, and FOV 50 mm. Datasets were reconstructed using 4 different kernels (B30f-smooth, B36f-medium smooth, B46f medium, and B60f-sharp). Each scan was scored for the presence of plaques. Once a plaque was detected, the operator performed attenuation measurements (HU) in coronary lumen, oil, calcified and soft plaque tissue using the same settings in all datasets. The results were compared with T-test and correlated with Pearson's test. Results. Overall, 464 measurements were performed. Significant differences (p [it

  19. Prediction model of velocity field around circular cylinder over various Reynolds numbers by fusion convolutional neural networks based on pressure on the cylinder

    Science.gov (United States)

    Jin, Xiaowei; Cheng, Peng; Chen, Wen-Li; Li, Hui

    2018-04-01

    A data-driven model is proposed for the prediction of the velocity field around a cylinder by fusion convolutional neural networks (CNNs) using measurements of the pressure field on the cylinder. The model is based on the close relationship between the Reynolds stresses in the wake, the wake formation length, and the base pressure. Numerical simulations of flow around a cylinder at various Reynolds numbers are carried out to establish a dataset capturing the effect of the Reynolds number on various flow properties. The time series of pressure fluctuations on the cylinder is converted into a grid-like spatial-temporal topology to be handled as the input of a CNN. A CNN architecture composed of a fusion of paths with and without a pooling layer is designed. This architecture can capture both accurate spatial-temporal information and the features that are invariant of small translations in the temporal dimension of pressure fluctuations on the cylinder. The CNN is trained using the computational fluid dynamics (CFD) dataset to establish the mapping relationship between the pressure fluctuations on the cylinder and the velocity field around the cylinder. Adam (adaptive moment estimation), an efficient method for processing large-scale and high-dimensional machine learning problems, is employed to implement the optimization algorithm. The trained model is then tested over various Reynolds numbers. The predictions of this model are found to agree well with the CFD results, and the data-driven model successfully learns the underlying flow regimes, i.e., the relationship between wake structure and pressure experienced on the surface of a cylinder is well established.

  20. Modeling and possible implementation of self-learning equivalence-convolutional neural structures for auto-encoding-decoding and clusterization of images

    Science.gov (United States)

    Krasilenko, Vladimir G.; Lazarev, Alexander A.; Nikitovich, Diana V.

    2017-08-01

    Self-learning equivalent-convolutional neural structures (SLECNS) for auto-coding-decoding and image clustering are discussed. The SLECNS architectures and their spatially invariant equivalent models (SI EMs) using the corresponding matrix-matrix procedures with basic operations of continuous logic and non-linear processing are proposed. These SI EMs have several advantages, such as the ability to recognize image fragments with better efficiency and strong cross correlation. The proposed clustering method of fragments with regard to their structural features is suitable not only for binary, but also color images and combines self-learning and the formation of weight clustered matrix-patterns. Its model is constructed and designed on the basis of recursively processing algorithms and to k-average method. The experimental results confirmed that larger images and 2D binary fragments with a large numbers of elements may be clustered. For the first time the possibility of generalization of these models for space invariant case is shown. The experiment for an image with dimension of 256x256 (a reference array) and fragments with dimensions of 7x7 and 21x21 for clustering is carried out. The experiments, using the software environment Mathcad, showed that the proposed method is universal, has a significant convergence, the small number of iterations is easily, displayed on the matrix structure, and confirmed its prospects. Thus, to understand the mechanisms of self-learning equivalence-convolutional clustering, accompanying her to the competitive processes in neurons, and the neural auto-encoding-decoding and recognition principles with the use of self-learning cluster patterns is very important which used the algorithm and the principles of non-linear processing of two-dimensional spatial functions of images comparison. These SIEMs can simply describe the signals processing during the all training and recognition stages and they are suitable for unipolar

  1. Modeling of biologically motivated self-learning equivalent-convolutional recurrent-multilayer neural structures (BLM_SL_EC_RMNS) for image fragments clustering and recognition

    Science.gov (United States)

    Krasilenko, Vladimir G.; Lazarev, Alexander A.; Nikitovich, Diana V.

    2018-03-01

    The biologically-motivated self-learning equivalence-convolutional recurrent-multilayer neural structures (BLM_SL_EC_RMNS) for fragments images clustering and recognition will be discussed. We shall consider these neural structures and their spatial-invariant equivalental models (SIEMs) based on proposed equivalent two-dimensional functions of image similarity and the corresponding matrix-matrix (or tensor) procedures using as basic operations of continuous logic and nonlinear processing. These SIEMs can simply describe the signals processing during the all training and recognition stages and they are suitable for unipolar-coding multilevel signals. The clustering efficiency in such models and their implementation depends on the discriminant properties of neural elements of hidden layers. Therefore, the main models and architecture parameters and characteristics depends on the applied types of non-linear processing and function used for image comparison or for adaptive-equivalent weighing of input patterns. We show that these SL_EC_RMNSs have several advantages, such as the self-study and self-identification of features and signs of the similarity of fragments, ability to clustering and recognize of image fragments with best efficiency and strong mutual correlation. The proposed combined with learning-recognition clustering method of fragments with regard to their structural features is suitable not only for binary, but also color images and combines self-learning and the formation of weight clustered matrix-patterns. Its model is constructed and designed on the basis of recursively continuous logic and nonlinear processing algorithms and to k-average method or method the winner takes all (WTA). The experimental results confirmed that fragments with a large numbers of elements may be clustered. For the first time the possibility of generalization of these models for space invariant case is shown. The experiment for an images of different dimensions (a reference

  2. DCMDN: Deep Convolutional Mixture Density Network

    Science.gov (United States)

    D'Isanto, Antonio; Polsterer, Kai Lars

    2017-09-01

    Deep Convolutional Mixture Density Network (DCMDN) estimates probabilistic photometric redshift directly from multi-band imaging data by combining a version of a deep convolutional network with a mixture density network. The estimates are expressed as Gaussian mixture models representing the probability density functions (PDFs) in the redshift space. In addition to the traditional scores, the continuous ranked probability score (CRPS) and the probability integral transform (PIT) are applied as performance criteria. DCMDN is able to predict redshift PDFs independently from the type of source, e.g. galaxies, quasars or stars and renders pre-classification of objects and feature extraction unnecessary; the method is extremely general and allows the solving of any kind of probabilistic regression problems based on imaging data, such as estimating metallicity or star formation rate in galaxies.

  3. Gas Classification Using Deep Convolutional Neural Networks

    Science.gov (United States)

    Peng, Pai; Zhao, Xiaojin; Pan, Xiaofang; Ye, Wenbin

    2018-01-01

    In this work, we propose a novel Deep Convolutional Neural Network (DCNN) tailored for gas classification. Inspired by the great success of DCNN in the field of computer vision, we designed a DCNN with up to 38 layers. In general, the proposed gas neural network, named GasNet, consists of: six convolutional blocks, each block consist of six layers; a pooling layer; and a fully-connected layer. Together, these various layers make up a powerful deep model for gas classification. Experimental results show that the proposed DCNN method is an effective technique for classifying electronic nose data. We also demonstrate that the DCNN method can provide higher classification accuracy than comparable Support Vector Machine (SVM) methods and Multiple Layer Perceptron (MLP). PMID:29316723

  4. Gas Classification Using Deep Convolutional Neural Networks.

    Science.gov (United States)

    Peng, Pai; Zhao, Xiaojin; Pan, Xiaofang; Ye, Wenbin

    2018-01-08

    In this work, we propose a novel Deep Convolutional Neural Network (DCNN) tailored for gas classification. Inspired by the great success of DCNN in the field of computer vision, we designed a DCNN with up to 38 layers. In general, the proposed gas neural network, named GasNet, consists of: six convolutional blocks, each block consist of six layers; a pooling layer; and a fully-connected layer. Together, these various layers make up a powerful deep model for gas classification. Experimental results show that the proposed DCNN method is an effective technique for classifying electronic nose data. We also demonstrate that the DCNN method can provide higher classification accuracy than comparable Support Vector Machine (SVM) methods and Multiple Layer Perceptron (MLP).

  5. Radial Structure Scaffolds Convolution Patterns of Developing Cerebral Cortex

    Directory of Open Access Journals (Sweden)

    Mir Jalil Razavi

    2017-08-01

    Full Text Available Commonly-preserved radial convolution is a prominent characteristic of the mammalian cerebral cortex. Endeavors from multiple disciplines have been devoted for decades to explore the causes for this enigmatic structure. However, the underlying mechanisms that lead to consistent cortical convolution patterns still remain poorly understood. In this work, inspired by prior studies, we propose and evaluate a plausible theory that radial convolution during the early development of the brain is sculptured by radial structures consisting of radial glial cells (RGCs and maturing axons. Specifically, the regionally heterogeneous development and distribution of RGCs controlled by Trnp1 regulate the convex and concave convolution patterns (gyri and sulci in the radial direction, while the interplay of RGCs' effects on convolution and axons regulates the convex (gyral convolution patterns. This theory is assessed by observations and measurements in literature from multiple disciplines such as neurobiology, genetics, biomechanics, etc., at multiple scales to date. Particularly, this theory is further validated by multimodal imaging data analysis and computational simulations in this study. We offer a versatile and descriptive study model that can provide reasonable explanations of observations, experiments, and simulations of the characteristic mammalian cortical folding.

  6. Video Super-Resolution via Bidirectional Recurrent Convolutional Networks.

    Science.gov (United States)

    Huang, Yan; Wang, Wei; Wang, Liang

    2018-04-01

    Super resolving a low-resolution video, namely video super-resolution (SR), is usually handled by either single-image SR or multi-frame SR. Single-Image SR deals with each video frame independently, and ignores intrinsic temporal dependency of video frames which actually plays a very important role in video SR. Multi-Frame SR generally extracts motion information, e.g., optical flow, to model the temporal dependency, but often shows high computational cost. Considering that recurrent neural networks (RNNs) can model long-term temporal dependency of video sequences well, we propose a fully convolutional RNN named bidirectional recurrent convolutional network for efficient multi-frame SR. Different from vanilla RNNs, 1) the commonly-used full feedforward and recurrent connections are replaced with weight-sharing convolutional connections. So they can greatly reduce the large number of network parameters and well model the temporal dependency in a finer level, i.e., patch-based rather than frame-based, and 2) connections from input layers at previous timesteps to the current hidden layer are added by 3D feedforward convolutions, which aim to capture discriminate spatio-temporal patterns for short-term fast-varying motions in local adjacent frames. Due to the cheap convolutional operations, our model has a low computational complexity and runs orders of magnitude faster than other multi-frame SR methods. With the powerful temporal dependency modeling, our model can super resolve videos with complex motions and achieve well performance.

  7. DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs.

    Science.gov (United States)

    Chen, Liang-Chieh; Papandreou, George; Kokkinos, Iasonas; Murphy, Kevin; Yuille, Alan L

    2018-04-01

    In this work we address the task of semantic image segmentation with Deep Learning and make three main contributions that are experimentally shown to have substantial practical merit. First, we highlight convolution with upsampled filters, or 'atrous convolution', as a powerful tool in dense prediction tasks. Atrous convolution allows us to explicitly control the resolution at which feature responses are computed within Deep Convolutional Neural Networks. It also allows us to effectively enlarge the field of view of filters to incorporate larger context without increasing the number of parameters or the amount of computation. Second, we propose atrous spatial pyramid pooling (ASPP) to robustly segment objects at multiple scales. ASPP probes an incoming convolutional feature layer with filters at multiple sampling rates and effective fields-of-views, thus capturing objects as well as image context at multiple scales. Third, we improve the localization of object boundaries by combining methods from DCNNs and probabilistic graphical models. The commonly deployed combination of max-pooling and downsampling in DCNNs achieves invariance but has a toll on localization accuracy. We overcome this by combining the responses at the final DCNN layer with a fully connected Conditional Random Field (CRF), which is shown both qualitatively and quantitatively to improve localization performance. Our proposed "DeepLab" system sets the new state-of-art at the PASCAL VOC-2012 semantic image segmentation task, reaching 79.7 percent mIOU in the test set, and advances the results on three other datasets: PASCAL-Context, PASCAL-Person-Part, and Cityscapes. All of our code is made publicly available online.

  8. Design of convolutional tornado code

    Science.gov (United States)

    Zhou, Hui; Yang, Yao; Gao, Hongmin; Tan, Lu

    2017-09-01

    As a linear block code, the traditional tornado (tTN) code is inefficient in burst-erasure environment and its multi-level structure may lead to high encoding/decoding complexity. This paper presents a convolutional tornado (cTN) code which is able to improve the burst-erasure protection capability by applying the convolution property to the tTN code, and reduce computational complexity by abrogating the multi-level structure. The simulation results show that cTN code can provide a better packet loss protection performance with lower computation complexity than tTN code.

  9. Detecting atrial fibrillation by deep convolutional neural networks.

    Science.gov (United States)

    Xia, Yong; Wulan, Naren; Wang, Kuanquan; Zhang, Henggui

    2018-02-01

    Atrial fibrillation (AF) is the most common cardiac arrhythmia. The incidence of AF increases with age, causing high risks of stroke and increased morbidity and mortality. Efficient and accurate diagnosis of AF based on the ECG is valuable in clinical settings and remains challenging. In this paper, we proposed a novel method with high reliability and accuracy for AF detection via deep learning. The short-term Fourier transform (STFT) and stationary wavelet transform (SWT) were used to analyze ECG segments to obtain two-dimensional (2-D) matrix input suitable for deep convolutional neural networks. Then, two different deep convolutional neural network models corresponding to STFT output and SWT output were developed. Our new method did not require detection of P or R peaks, nor feature designs for classification, in contrast to existing algorithms. Finally, the performances of the two models were evaluated and compared with those of existing algorithms. Our proposed method demonstrated favorable performances on ECG segments as short as 5 s. The deep convolutional neural network using input generated by STFT, presented a sensitivity of 98.34%, specificity of 98.24% and accuracy of 98.29%. For the deep convolutional neural network using input generated by SWT, a sensitivity of 98.79%, specificity of 97.87% and accuracy of 98.63% was achieved. The proposed method using deep convolutional neural networks shows high sensitivity, specificity and accuracy, and, therefore, is a valuable tool for AF detection. Copyright © 2017 Elsevier Ltd. All rights reserved.

  10. Solutions to Arithmetic Convolution Equations

    Czech Academy of Sciences Publication Activity Database

    Glöckner, H.; Lucht, L.G.; Porubský, Štefan

    2007-01-01

    Roč. 135, č. 6 (2007), s. 1619-1629 ISSN 0002-9939 R&D Projects: GA ČR GA201/04/0381 Institutional research plan: CEZ:AV0Z10300504 Keywords : arithmetic functions * Dirichlet convolution * polynomial equations * analytic equations * topological algebras * holomorphic functional calculus Subject RIV: BA - General Mathematics Impact factor: 0.520, year: 2007

  11. Decoding LDPC Convolutional Codes on Markov Channels

    Directory of Open Access Journals (Sweden)

    Kashyap Manohar

    2008-01-01

    Full Text Available Abstract This paper describes a pipelined iterative technique for joint decoding and channel state estimation of LDPC convolutional codes over Markov channels. Example designs are presented for the Gilbert-Elliott discrete channel model. We also compare the performance and complexity of our algorithm against joint decoding and state estimation of conventional LDPC block codes. Complexity analysis reveals that our pipelined algorithm reduces the number of operations per time step compared to LDPC block codes, at the expense of increased memory and latency. This tradeoff is favorable for low-power applications.

  12. Decoding LDPC Convolutional Codes on Markov Channels

    Directory of Open Access Journals (Sweden)

    Chris Winstead

    2008-04-01

    Full Text Available This paper describes a pipelined iterative technique for joint decoding and channel state estimation of LDPC convolutional codes over Markov channels. Example designs are presented for the Gilbert-Elliott discrete channel model. We also compare the performance and complexity of our algorithm against joint decoding and state estimation of conventional LDPC block codes. Complexity analysis reveals that our pipelined algorithm reduces the number of operations per time step compared to LDPC block codes, at the expense of increased memory and latency. This tradeoff is favorable for low-power applications.

  13. Adaptive Graph Convolutional Neural Networks

    OpenAIRE

    Li, Ruoyu; Wang, Sheng; Zhu, Feiyun; Huang, Junzhou

    2018-01-01

    Graph Convolutional Neural Networks (Graph CNNs) are generalizations of classical CNNs to handle graph data such as molecular data, point could and social networks. Current filters in graph CNNs are built for fixed and shared graph structure. However, for most real data, the graph structures varies in both size and connectivity. The paper proposes a generalized and flexible graph CNN taking data of arbitrary graph structure as input. In that way a task-driven adaptive graph is learned for eac...

  14. Review of the convolution algorithm for evaluating service integrated systems

    DEFF Research Database (Denmark)

    Iversen, Villy Bæk

    1997-01-01

    In this paper we give a review of the applicability of the convolution algorithm. By this we are able to evaluate communication networks end--to--end with e.g. BPP multi-ratetraffic models insensitive to the holding time distribution. Rearrangement, minimum allocation, and maximum allocation...

  15. Convolution of Distribution-Valued Functions. Applications.

    OpenAIRE

    BARGETZ, CHRISTIAN

    2011-01-01

    In this article we examine products and convolutions of vector-valued functions. For nuclear normal spaces of distributions Proposition 25 in [31,p. 120] yields a vector-valued product or convolution if there is a continuous product or convolution mapping in the range of the vector-valued functions. For specific spaces, we generalize this result to hypocontinuous bilinear maps at the expense of generality with respect to the function space. We consider holomorphic, meromorphic and differentia...

  16. Down image recognition based on deep convolutional neural network

    Directory of Open Access Journals (Sweden)

    Wenzhu Yang

    2018-06-01

    Full Text Available Since of the scale and the various shapes of down in the image, it is difficult for traditional image recognition method to correctly recognize the type of down image and get the required recognition accuracy, even for the Traditional Convolutional Neural Network (TCNN. To deal with the above problems, a Deep Convolutional Neural Network (DCNN for down image classification is constructed, and a new weight initialization method is proposed. Firstly, the salient regions of a down image were cut from the image using the visual saliency model. Then, these salient regions of the image were used to train a sparse autoencoder and get a collection of convolutional filters, which accord with the statistical characteristics of dataset. At last, a DCNN with Inception module and its variants was constructed. To improve the recognition accuracy, the depth of the network is deepened. The experiment results indicate that the constructed DCNN increases the recognition accuracy by 2.7% compared to TCNN, when recognizing the down in the images. The convergence rate of the proposed DCNN with the new weight initialization method is improved by 25.5% compared to TCNN. Keywords: Deep convolutional neural network, Weight initialization, Sparse autoencoder, Visual saliency model, Image recognition

  17. A convolutional neural network to filter artifacts in spectroscopic MRI.

    Science.gov (United States)

    Gurbani, Saumya S; Schreibmann, Eduard; Maudsley, Andrew A; Cordova, James Scott; Soher, Brian J; Poptani, Harish; Verma, Gaurav; Barker, Peter B; Shim, Hyunsuk; Cooper, Lee A D

    2018-03-09

    Proton MRSI is a noninvasive modality capable of generating volumetric maps of in vivo tissue metabolism without the need for ionizing radiation or injected contrast agent. Magnetic resonance spectroscopic imaging has been shown to be a viable imaging modality for studying several neuropathologies. However, a key hurdle in the routine clinical adoption of MRSI is the presence of spectral artifacts that can arise from a number of sources, possibly leading to false information. A deep learning model was developed that was capable of identifying and filtering out poor quality spectra. The core of the model used a tiled convolutional neural network that analyzed frequency-domain spectra to detect artifacts. When compared with a panel of MRS experts, our convolutional neural network achieved high sensitivity and specificity with an area under the curve of 0.95. A visualization scheme was implemented to better understand how the convolutional neural network made its judgement on single-voxel or multivoxel MRSI, and the convolutional neural network was embedded into a pipeline capable of producing whole-brain spectroscopic MRI volumes in real time. The fully automated method for assessment of spectral quality provides a valuable tool to support clinical MRSI or spectroscopic MRI studies for use in fields such as adaptive radiation therapy planning. © 2018 International Society for Magnetic Resonance in Medicine.

  18. Deep learning for steganalysis via convolutional neural networks

    Science.gov (United States)

    Qian, Yinlong; Dong, Jing; Wang, Wei; Tan, Tieniu

    2015-03-01

    Current work on steganalysis for digital images is focused on the construction of complex handcrafted features. This paper proposes a new paradigm for steganalysis to learn features automatically via deep learning models. We novelly propose a customized Convolutional Neural Network for steganalysis. The proposed model can capture the complex dependencies that are useful for steganalysis. Compared with existing schemes, this model can automatically learn feature representations with several convolutional layers. The feature extraction and classification steps are unified under a single architecture, which means the guidance of classification can be used during the feature extraction step. We demonstrate the effectiveness of the proposed model on three state-of-theart spatial domain steganographic algorithms - HUGO, WOW, and S-UNIWARD. Compared to the Spatial Rich Model (SRM), our model achieves comparable performance on BOSSbase and the realistic and large ImageNet database.

  19. Learning Convolutional Text Representations for Visual Question Answering

    OpenAIRE

    Wang, Zhengyang; Ji, Shuiwang

    2017-01-01

    Visual question answering is a recently proposed artificial intelligence task that requires a deep understanding of both images and texts. In deep learning, images are typically modeled through convolutional neural networks, and texts are typically modeled through recurrent neural networks. While the requirement for modeling images is similar to traditional computer vision tasks, such as object recognition and image classification, visual question answering raises a different need for textual...

  20. Maximum likelihood convolutional decoding (MCD) performance due to system losses

    Science.gov (United States)

    Webster, L.

    1976-01-01

    A model for predicting the computational performance of a maximum likelihood convolutional decoder (MCD) operating in a noisy carrier reference environment is described. This model is used to develop a subroutine that will be utilized by the Telemetry Analysis Program to compute the MCD bit error rate. When this computational model is averaged over noisy reference phase errors using a high-rate interpolation scheme, the results are found to agree quite favorably with experimental measurements.

  1. A Study of Recurrent and Convolutional Neural Networks in the Native Language Identification Task

    KAUST Repository

    Werfelmann, Robert

    2018-01-01

    around the world. The neural network models consisted of Long Short-Term Memory and Convolutional networks using the sentences of each document as the input. Additional statistical features were generated from the text to complement the predictions

  2. Symbol synchronization in convolutionally coded systems

    Science.gov (United States)

    Baumert, L. D.; Mceliece, R. J.; Van Tilborg, H. C. A.

    1979-01-01

    Alternate symbol inversion is sometimes applied to the output of convolutional encoders to guarantee sufficient richness of symbol transition for the receiver symbol synchronizer. A bound is given for the length of the transition-free symbol stream in such systems, and those convolutional codes are characterized in which arbitrarily long transition free runs occur.

  3. The general theory of convolutional codes

    Science.gov (United States)

    Mceliece, R. J.; Stanley, R. P.

    1993-01-01

    This article presents a self-contained introduction to the algebraic theory of convolutional codes. This introduction is partly a tutorial, but at the same time contains a number of new results which will prove useful for designers of advanced telecommunication systems. Among the new concepts introduced here are the Hilbert series for a convolutional code and the class of compact codes.

  4. Convolutions

    Indian Academy of Sciences (India)

    President's Address to the Association of Mathematics Teachers of India, December 2011. I am expected to tell you, in 25 minutes, something that should interest you, excite you, pique your curiosity, and make you look for more. It is a tall order, but I will try. The word 'interactive' is in fashion these days. So I will leave a few ...

  5. Improving deep convolutional neural networks with mixed maxout units.

    Directory of Open Access Journals (Sweden)

    Hui-Zhen Zhao

    Full Text Available Motivated by insights from the maxout-units-based deep Convolutional Neural Network (CNN that "non-maximal features are unable to deliver" and "feature mapping subspace pooling is insufficient," we present a novel mixed variant of the recently introduced maxout unit called a mixout unit. Specifically, we do so by calculating the exponential probabilities of feature mappings gained by applying different convolutional transformations over the same input and then calculating the expected values according to their exponential probabilities. Moreover, we introduce the Bernoulli distribution to balance the maximum values with the expected values of the feature mappings subspace. Finally, we design a simple model to verify the pooling ability of mixout units and a Mixout-units-based Network-in-Network (NiN model to analyze the feature learning ability of the mixout models. We argue that our proposed units improve the pooling ability and that mixout models can achieve better feature learning and classification performance.

  6. Resting State fMRI Functional Connectivity-Based Classification Using a Convolutional Neural Network Architecture.

    Science.gov (United States)

    Meszlényi, Regina J; Buza, Krisztian; Vidnyánszky, Zoltán

    2017-01-01

    Machine learning techniques have become increasingly popular in the field of resting state fMRI (functional magnetic resonance imaging) network based classification. However, the application of convolutional networks has been proposed only very recently and has remained largely unexplored. In this paper we describe a convolutional neural network architecture for functional connectome classification called connectome-convolutional neural network (CCNN). Our results on simulated datasets and a publicly available dataset for amnestic mild cognitive impairment classification demonstrate that our CCNN model can efficiently distinguish between subject groups. We also show that the connectome-convolutional network is capable to combine information from diverse functional connectivity metrics and that models using a combination of different connectivity descriptors are able to outperform classifiers using only one metric. From this flexibility follows that our proposed CCNN model can be easily adapted to a wide range of connectome based classification or regression tasks, by varying which connectivity descriptor combinations are used to train the network.

  7. Adaptive decoding of convolutional codes

    Directory of Open Access Journals (Sweden)

    K. Hueske

    2007-06-01

    Full Text Available Convolutional codes, which are frequently used as error correction codes in digital transmission systems, are generally decoded using the Viterbi Decoder. On the one hand the Viterbi Decoder is an optimum maximum likelihood decoder, i.e. the most probable transmitted code sequence is obtained. On the other hand the mathematical complexity of the algorithm only depends on the used code, not on the number of transmission errors. To reduce the complexity of the decoding process for good transmission conditions, an alternative syndrome based decoder is presented. The reduction of complexity is realized by two different approaches, the syndrome zero sequence deactivation and the path metric equalization. The two approaches enable an easy adaptation of the decoding complexity for different transmission conditions, which results in a trade-off between decoding complexity and error correction performance.

  8. Adaptive decoding of convolutional codes

    Science.gov (United States)

    Hueske, K.; Geldmacher, J.; Götze, J.

    2007-06-01

    Convolutional codes, which are frequently used as error correction codes in digital transmission systems, are generally decoded using the Viterbi Decoder. On the one hand the Viterbi Decoder is an optimum maximum likelihood decoder, i.e. the most probable transmitted code sequence is obtained. On the other hand the mathematical complexity of the algorithm only depends on the used code, not on the number of transmission errors. To reduce the complexity of the decoding process for good transmission conditions, an alternative syndrome based decoder is presented. The reduction of complexity is realized by two different approaches, the syndrome zero sequence deactivation and the path metric equalization. The two approaches enable an easy adaptation of the decoding complexity for different transmission conditions, which results in a trade-off between decoding complexity and error correction performance.

  9. Application of structured support vector machine backpropagation to a convolutional neural network for human pose estimation.

    Science.gov (United States)

    Witoonchart, Peerajak; Chongstitvatana, Prabhas

    2017-08-01

    In this study, for the first time, we show how to formulate a structured support vector machine (SSVM) as two layers in a convolutional neural network, where the top layer is a loss augmented inference layer and the bottom layer is the normal convolutional layer. We show that a deformable part model can be learned with the proposed structured SVM neural network by backpropagating the error of the deformable part model to the convolutional neural network. The forward propagation calculates the loss augmented inference and the backpropagation calculates the gradient from the loss augmented inference layer to the convolutional layer. Thus, we obtain a new type of convolutional neural network called an Structured SVM convolutional neural network, which we applied to the human pose estimation problem. This new neural network can be used as the final layers in deep learning. Our method jointly learns the structural model parameters and the appearance model parameters. We implemented our method as a new layer in the existing Caffe library. Copyright © 2017 Elsevier Ltd. All rights reserved.

  10. A Convolution Tree with Deconvolution Branches: Exploiting Geometric Relationships for Single Shot Keypoint Detection

    OpenAIRE

    Kumar, Amit; Chellappa, Rama

    2017-01-01

    Recently, Deep Convolution Networks (DCNNs) have been applied to the task of face alignment and have shown potential for learning improved feature representations. Although deeper layers can capture abstract concepts like pose, it is difficult to capture the geometric relationships among the keypoints in DCNNs. In this paper, we propose a novel convolution-deconvolution network for facial keypoint detection. Our model predicts the 2D locations of the keypoints and their individual visibility ...

  11. Learning text representation using recurrent convolutional neural network with highway layers

    OpenAIRE

    Wen, Ying; Zhang, Weinan; Luo, Rui; Wang, Jun

    2016-01-01

    Recently, the rapid development of word embedding and neural networks has brought new inspiration to various NLP and IR tasks. In this paper, we describe a staged hybrid model combining Recurrent Convolutional Neural Networks (RCNN) with highway layers. The highway network module is incorporated in the middle takes the output of the bi-directional Recurrent Neural Network (Bi-RNN) module in the first stage and provides the Convolutional Neural Network (CNN) module in the last stage with the i...

  12. Segmentation of Drosophila Heart in Optical Coherence Microscopy Images Using Convolutional Neural Networks

    OpenAIRE

    Duan, Lian; Qin, Xi; He, Yuanhao; Sang, Xialin; Pan, Jinda; Xu, Tao; Men, Jing; Tanzi, Rudolph E.; Li, Airong; Ma, Yutao; Zhou, Chao

    2018-01-01

    Convolutional neural networks are powerful tools for image segmentation and classification. Here, we use this method to identify and mark the heart region of Drosophila at different developmental stages in the cross-sectional images acquired by a custom optical coherence microscopy (OCM) system. With our well-trained convolutional neural network model, the heart regions through multiple heartbeat cycles can be marked with an intersection over union (IOU) of ~86%. Various morphological and dyn...

  13. Convolutional Neural Network for Image Recognition

    CERN Document Server

    Seifnashri, Sahand

    2015-01-01

    The aim of this project is to use machine learning techniques especially Convolutional Neural Networks for image processing. These techniques can be used for Quark-Gluon discrimination using calorimeters data, but unfortunately I didn’t manage to get the calorimeters data and I just used the Jet data fromminiaodsim(ak4 chs). The Jet data was not good enough for Convolutional Neural Network which is designed for ’image’ recognition. This report is made of twomain part, part one is mainly about implementing Convolutional Neural Network on unphysical data such as MNIST digits and CIFAR-10 dataset and part 2 is about the Jet data.

  14. Infimal Convolution Regularisation Functionals of BV and Lp Spaces

    KAUST Repository

    Burger, Martin

    2016-02-03

    We study a general class of infimal convolution type regularisation functionals suitable for applications in image processing. These functionals incorporate a combination of the total variation seminorm and Lp norms. A unified well-posedness analysis is presented and a detailed study of the one-dimensional model is performed, by computing exact solutions for the corresponding denoising problem and the case p=2. Furthermore, the dependency of the regularisation properties of this infimal convolution approach to the choice of p is studied. It turns out that in the case p=2 this regulariser is equivalent to the Huber-type variant of total variation regularisation. We provide numerical examples for image decomposition as well as for image denoising. We show that our model is capable of eliminating the staircasing effect, a well-known disadvantage of total variation regularisation. Moreover as p increases we obtain almost piecewise affine reconstructions, leading also to a better preservation of hat-like structures.

  15. Convolution of large 3D images on GPU and its decomposition

    Science.gov (United States)

    Karas, Pavel; Svoboda, David

    2011-12-01

    In this article, we propose a method for computing convolution of large 3D images. The convolution is performed in a frequency domain using a convolution theorem. The algorithm is accelerated on a graphic card by means of the CUDA parallel computing model. Convolution is decomposed in a frequency domain using the decimation in frequency algorithm. We pay attention to keeping our approach efficient in terms of both time and memory consumption and also in terms of memory transfers between CPU and GPU which have a significant inuence on overall computational time. We also study the implementation on multiple GPUs and compare the results between the multi-GPU and multi-CPU implementations.

  16. Multiscale Convolutional Neural Networks for Hand Detection

    Directory of Open Access Journals (Sweden)

    Shiyang Yan

    2017-01-01

    Full Text Available Unconstrained hand detection in still images plays an important role in many hand-related vision problems, for example, hand tracking, gesture analysis, human action recognition and human-machine interaction, and sign language recognition. Although hand detection has been extensively studied for decades, it is still a challenging task with many problems to be tackled. The contributing factors for this complexity include heavy occlusion, low resolution, varying illumination conditions, different hand gestures, and the complex interactions between hands and objects or other hands. In this paper, we propose a multiscale deep learning model for unconstrained hand detection in still images. Deep learning models, and deep convolutional neural networks (CNNs in particular, have achieved state-of-the-art performances in many vision benchmarks. Developed from the region-based CNN (R-CNN model, we propose a hand detection scheme based on candidate regions generated by a generic region proposal algorithm, followed by multiscale information fusion from the popular VGG16 model. Two benchmark datasets were applied to validate the proposed method, namely, the Oxford Hand Detection Dataset and the VIVA Hand Detection Challenge. We achieved state-of-the-art results on the Oxford Hand Detection Dataset and had satisfactory performance in the VIVA Hand Detection Challenge.

  17. A Note on Cubic Convolution Interpolation

    OpenAIRE

    Meijering, E.; Unser, M.

    2003-01-01

    We establish a link between classical osculatory interpolation and modern convolution-based interpolation and use it to show that two well-known cubic convolution schemes are formally equivalent to two osculatory interpolation schemes proposed in the actuarial literature about a century ago. We also discuss computational differences and give examples of other cubic interpolation schemes not previously studied in signal and image processing.

  18. Convolutional over Recurrent Encoder for Neural Machine Translation

    Directory of Open Access Journals (Sweden)

    Dakwale Praveen

    2017-06-01

    Full Text Available Neural machine translation is a recently proposed approach which has shown competitive results to traditional MT approaches. Standard neural MT is an end-to-end neural network where the source sentence is encoded by a recurrent neural network (RNN called encoder and the target words are predicted using another RNN known as decoder. Recently, various models have been proposed which replace the RNN encoder with a convolutional neural network (CNN. In this paper, we propose to augment the standard RNN encoder in NMT with additional convolutional layers in order to capture wider context in the encoder output. Experiments on English to German translation demonstrate that our approach can achieve significant improvements over a standard RNN-based baseline.

  19. A Fast Numerical Method for Max-Convolution and the Application to Efficient Max-Product Inference in Bayesian Networks.

    Science.gov (United States)

    Serang, Oliver

    2015-08-01

    Observations depending on sums of random variables are common throughout many fields; however, no efficient solution is currently known for performing max-product inference on these sums of general discrete distributions (max-product inference can be used to obtain maximum a posteriori estimates). The limiting step to max-product inference is the max-convolution problem (sometimes presented in log-transformed form and denoted as "infimal convolution," "min-convolution," or "convolution on the tropical semiring"), for which no O(k log(k)) method is currently known. Presented here is an O(k log(k)) numerical method for estimating the max-convolution of two nonnegative vectors (e.g., two probability mass functions), where k is the length of the larger vector. This numerical max-convolution method is then demonstrated by performing fast max-product inference on a convolution tree, a data structure for performing fast inference given information on the sum of n discrete random variables in O(nk log(nk)log(n)) steps (where each random variable has an arbitrary prior distribution on k contiguous possible states). The numerical max-convolution method can be applied to specialized classes of hidden Markov models to reduce the runtime of computing the Viterbi path from nk(2) to nk log(k), and has potential application to the all-pairs shortest paths problem.

  20. AFM tip-sample convolution effects for cylinder protrusions

    Science.gov (United States)

    Shen, Jian; Zhang, Dan; Zhang, Fei-Hu; Gan, Yang

    2017-11-01

    A thorough understanding about the AFM tip geometry dependent artifacts and tip-sample convolution effect is essential for reliable AFM topographic characterization and dimensional metrology. Using rigid sapphire cylinder protrusions (diameter: 2.25 μm, height: 575 nm) as the model system, a systematic and quantitative study about the imaging artifacts of four types of tips-two different pyramidal tips, one tetrahedral tip and one super sharp whisker tip-is carried out through comparing tip geometry dependent variations in AFM topography of cylinders and constructing the rigid tip-cylinder convolution models. We found that the imaging artifacts and the tip-sample convolution effect are critically related to the actual inclination of the working cantilever, the tip geometry, and the obstructive contacts between the working tip's planes/edges and the cylinder. Artifact-free images can only be obtained provided that all planes and edges of the working tip are steeper than the cylinder sidewalls. The findings reported here will contribute to reliable AFM characterization of surface features of micron or hundreds of nanometers in height that are frequently met in semiconductor, biology and materials fields.

  1. Nuclear norm regularized convolutional Max Pos@Top machine

    KAUST Repository

    Li, Qinfeng

    2016-11-18

    In this paper, we propose a novel classification model for the multiple instance data, which aims to maximize the number of positive instances ranked before the top-ranked negative instances. This method belongs to a recently emerged performance, named as Pos@Top. Our proposed classification model has a convolutional structure that is composed by four layers, i.e., the convolutional layer, the activation layer, the max-pooling layer and the full connection layer. In this paper, we propose an algorithm to learn the convolutional filters and the full connection weights to maximize the Pos@Top measure over the training set. Also, we try to minimize the rank of the filter matrix to explore the low-dimensional space of the instances in conjunction with the classification results. The rank minimization is conducted by the nuclear norm minimization of the filter matrix. In addition, we develop an iterative algorithm to solve the corresponding problem. We test our method on several benchmark datasets. The experimental results show the superiority of our method compared with other state-of-the-art Pos@Top maximization methods.

  2. Salient regions detection using convolutional neural networks and color volume

    Science.gov (United States)

    Liu, Guang-Hai; Hou, Yingkun

    2018-03-01

    Convolutional neural network is an important technique in machine learning, pattern recognition and image processing. In order to reduce the computational burden and extend the classical LeNet-5 model to the field of saliency detection, we propose a simple and novel computing model based on LeNet-5 network. In the proposed model, hue, saturation and intensity are utilized to extract depth cues, and then we integrate depth cues and color volume to saliency detection following the basic structure of the feature integration theory. Experimental results show that the proposed computing model outperforms some existing state-of-the-art methods on MSRA1000 and ECSSD datasets.

  3. Classifying medical relations in clinical text via convolutional neural networks.

    Science.gov (United States)

    He, Bin; Guan, Yi; Dai, Rui

    2018-05-16

    Deep learning research on relation classification has achieved solid performance in the general domain. This study proposes a convolutional neural network (CNN) architecture with a multi-pooling operation for medical relation classification on clinical records and explores a loss function with a category-level constraint matrix. Experiments using the 2010 i2b2/VA relation corpus demonstrate these models, which do not depend on any external features, outperform previous single-model methods and our best model is competitive with the existing ensemble-based method. Copyright © 2018. Published by Elsevier B.V.

  4. Do Convolutional Neural Networks Learn Class Hierarchy?

    Science.gov (United States)

    Bilal, Alsallakh; Jourabloo, Amin; Ye, Mao; Liu, Xiaoming; Ren, Liu

    2018-01-01

    Convolutional Neural Networks (CNNs) currently achieve state-of-the-art accuracy in image classification. With a growing number of classes, the accuracy usually drops as the possibilities of confusion increase. Interestingly, the class confusion patterns follow a hierarchical structure over the classes. We present visual-analytics methods to reveal and analyze this hierarchy of similar classes in relation with CNN-internal data. We found that this hierarchy not only dictates the confusion patterns between the classes, it furthermore dictates the learning behavior of CNNs. In particular, the early layers in these networks develop feature detectors that can separate high-level groups of classes quite well, even after a few training epochs. In contrast, the latter layers require substantially more epochs to develop specialized feature detectors that can separate individual classes. We demonstrate how these insights are key to significant improvement in accuracy by designing hierarchy-aware CNNs that accelerate model convergence and alleviate overfitting. We further demonstrate how our methods help in identifying various quality issues in the training data.

  5. Traffic sign classification with dataset augmentation and convolutional neural network

    Science.gov (United States)

    Tang, Qing; Kurnianggoro, Laksono; Jo, Kang-Hyun

    2018-04-01

    This paper presents a method for traffic sign classification using a convolutional neural network (CNN). In this method, firstly we transfer a color image into grayscale, and then normalize it in the range (-1,1) as the preprocessing step. To increase robustness of classification model, we apply a dataset augmentation algorithm and create new images to train the model. To avoid overfitting, we utilize a dropout module before the last fully connection layer. To assess the performance of the proposed method, the German traffic sign recognition benchmark (GTSRB) dataset is utilized. Experimental results show that the method is effective in classifying traffic signs.

  6. Semantic segmentation of bioimages using convolutional neural networks

    CSIR Research Space (South Africa)

    Wiehman, S

    2016-07-01

    Full Text Available Convolutional neural networks have shown great promise in both general image segmentation problems as well as bioimage segmentation. In this paper, the application of different convolutional network architectures is explored on the C. elegans live...

  7. One weird trick for parallelizing convolutional neural networks

    OpenAIRE

    Krizhevsky, Alex

    2014-01-01

    I present a new way to parallelize the training of convolutional neural networks across multiple GPUs. The method scales significantly better than all alternatives when applied to modern convolutional neural networks.

  8. Gradient Flow Convolutive Blind Source Separation

    DEFF Research Database (Denmark)

    Pedersen, Michael Syskind; Nielsen, Chinton Møller

    2004-01-01

    Experiments have shown that the performance of instantaneous gradient flow beamforming by Cauwenberghs et al. is reduced significantly in reverberant conditions. By expanding the gradient flow principle to convolutive mixtures, separation in a reverberant environment is possible. By use...... of a circular four microphone array with a radius of 5 mm, and applying convolutive gradient flow instead of just applying instantaneous gradient flow, experimental results show an improvement of up to around 14 dB can be achieved for simulated impulse responses and up to around 10 dB for a hearing aid...

  9. Human Face Recognition Using Convolutional Neural Networks

    Directory of Open Access Journals (Sweden)

    Răzvan-Daniel Albu

    2009-10-01

    Full Text Available In this paper, I present a novel hybrid face recognition approach based on a convolutional neural architecture, designed to robustly detect highly variable face patterns. The convolutional network extracts successively larger features in a hierarchical set of layers. With the weights of the trained neural networks there are created kernel windows used for feature extraction in a 3-stage algorithm. I present experimental results illustrating the efficiency of the proposed approach. I use a database of 796 images of 159 individuals from Reims University which contains quite a high degree of variability in expression, pose, and facial details.

  10. CMOS Compressed Imaging by Random Convolution

    OpenAIRE

    Jacques, Laurent; Vandergheynst, Pierre; Bibet, Alexandre; Majidzadeh, Vahid; Schmid, Alexandre; Leblebici, Yusuf

    2009-01-01

    We present a CMOS imager with built-in capability to perform Compressed Sensing. The adopted sensing strategy is the random Convolution due to J. Romberg. It is achieved by a shift register set in a pseudo-random configuration. It acts as a convolutive filter on the imager focal plane, the current issued from each CMOS pixel undergoing a pseudo-random redirection controlled by each component of the filter sequence. A pseudo-random triggering of the ADC reading is finally applied to comp...

  11. Feedback equivalence of convolutional codes over finite rings

    Directory of Open Access Journals (Sweden)

    DeCastro-García Noemí

    2017-12-01

    Full Text Available The approach to convolutional codes from the linear systems point of view provides us with effective tools in order to construct convolutional codes with adequate properties that let us use them in many applications. In this work, we have generalized feedback equivalence between families of convolutional codes and linear systems over certain rings, and we show that every locally Brunovsky linear system may be considered as a representation of a code under feedback convolutional equivalence.

  12. Efficient airport detection using region-based fully convolutional neural networks

    Science.gov (United States)

    Xin, Peng; Xu, Yuelei; Zhang, Xulei; Ma, Shiping; Li, Shuai; Lv, Chao

    2018-04-01

    This paper presents a model for airport detection using region-based fully convolutional neural networks. To achieve fast detection with high accuracy, we shared the conv layers between the region proposal procedure and the airport detection procedure and used graphics processing units (GPUs) to speed up the training and testing time. For lack of labeled data, we transferred the convolutional layers of ZF net pretrained by ImageNet to initialize the shared convolutional layers, then we retrained the model using the alternating optimization training strategy. The proposed model has been tested on an airport dataset consisting of 600 images. Experiments show that the proposed method can distinguish airports in our dataset from similar background scenes almost real-time with high accuracy, which is much better than traditional methods.

  13. Discrete convolution-operators and radioactive disintegration. [Numerical solution

    Energy Technology Data Exchange (ETDEWEB)

    Kalla, S L; VALENTINUZZI, M E [UNIVERSIDAD NACIONAL DE TUCUMAN (ARGENTINA). FACULTAD DE CIENCIAS EXACTAS Y TECNOLOGIA

    1975-08-01

    The basic concepts of discrete convolution and discrete convolution-operators are briefly described. Then, using the discrete convolution - operators, the differential equations associated with the process of radioactive disintegration are numerically solved. The importance of the method is emphasized to solve numerically, differential and integral equations.

  14. Deformable image registration using convolutional neural networks

    NARCIS (Netherlands)

    Eppenhof, Koen A.J.; Lafarge, Maxime W.; Moeskops, Pim; Veta, Mitko; Pluim, Josien P.W.

    2018-01-01

    Deformable image registration can be time-consuming and often needs extensive parameterization to perform well on a specific application. We present a step towards a registration framework based on a three-dimensional convolutional neural network. The network directly learns transformations between

  15. Epileptiform spike detection via convolutional neural networks

    DEFF Research Database (Denmark)

    Johansen, Alexander Rosenberg; Jin, Jing; Maszczyk, Tomasz

    2016-01-01

    The EEG of epileptic patients often contains sharp waveforms called "spikes", occurring between seizures. Detecting such spikes is crucial for diagnosing epilepsy. In this paper, we develop a convolutional neural network (CNN) for detecting spikes in EEG of epileptic patients in an automated...

  16. Convolutional Neural Networks for SAR Image Segmentation

    DEFF Research Database (Denmark)

    Malmgren-Hansen, David; Nobel-Jørgensen, Morten

    2015-01-01

    Segmentation of Synthetic Aperture Radar (SAR) images has several uses, but it is a difficult task due to a number of properties related to SAR images. In this article we show how Convolutional Neural Networks (CNNs) can easily be trained for SAR image segmentation with good results. Besides...

  17. A locality aware convolutional neural networks accelerator

    NARCIS (Netherlands)

    Shi, R.; Xu, Z.; Sun, Z.; Peemen, M.C.J.; Li, A.; Corporaal, H.; Wu, D.

    2015-01-01

    The advantages of Convolutional Neural Networks (CNNs) with respect to traditional methods for visual pattern recognition have changed the field of machine vision. The main issue that hinders broad adoption of this technique is the massive computing workload in CNN that prevents real-time

  18. Estimating the number of sources in a noisy convolutive mixture using BIC

    DEFF Research Database (Denmark)

    Olsson, Rasmus Kongsgaard; Hansen, Lars Kai

    2004-01-01

    The number of source signals in a noisy convolutive mixture is determined based on the exact log-likelihoods of the candidate models. In (Olsson and Hansen, 2004), a novel probabilistic blind source separator was introduced that is based solely on the time-varying second-order statistics of the s......The number of source signals in a noisy convolutive mixture is determined based on the exact log-likelihoods of the candidate models. In (Olsson and Hansen, 2004), a novel probabilistic blind source separator was introduced that is based solely on the time-varying second-order statistics...

  19. Alternate symbol inversion for improved symbol synchronization in convolutionally coded systems

    Science.gov (United States)

    Simon, M. K.; Smith, J. G.

    1980-01-01

    Inverting alternate symbols of the encoder output of a convolutionally coded system provides sufficient density of symbol transitions to guarantee adequate symbol synchronizer performance, a guarantee otherwise lacking. Although alternate symbol inversion may increase or decrease the average transition density, depending on the data source model, it produces a maximum number of contiguous symbols without transition for a particular class of convolutional codes, independent of the data source model. Further, this maximum is sufficiently small to guarantee acceptable symbol synchronizer performance for typical applications. Subsequent inversion of alternate detected symbols permits proper decoding.

  20. Convolute laminations — a theoretical analysis: example of a Pennsylvanian sandstone

    Science.gov (United States)

    Visher, Glenn S.; Cunningham, Russ D.

    1981-03-01

    Data from an outcropping laminated interval were collected and analyzed to test the applicability of a theoretical model describing instability of layered systems. Rayleigh—Taylor wave perturbations result at the interface between fluids of contrasting density, viscosity, and thickness. In the special case where reverse density and viscosity interlaminations are developed, the deformation response produces a single wave with predictable amplitudes, wavelengths, and amplification rates. Physical measurements from both the outcropping section and modern sediments suggest the usefulness of the model for the interpretation of convolute laminations. Internal characteristics of the stratigraphic interval, and the developmental sequence of convoluted beds, are used to document the developmental history of these structures.

  1. Design and Implementation of Behavior Recognition System Based on Convolutional Neural Network

    Directory of Open Access Journals (Sweden)

    Yu Bo

    2017-01-01

    Full Text Available We build a set of human behavior recognition system based on the convolution neural network constructed for the specific human behavior in public places. Firstly, video of human behavior data set will be segmented into images, then we process the images by the method of background subtraction to extract moving foreground characters of body. Secondly, the training data sets are trained into the designed convolution neural network, and the depth learning network is constructed by stochastic gradient descent. Finally, the various behaviors of samples are classified and identified with the obtained network model, and the recognition results are compared with the current mainstream methods. The result show that the convolution neural network can study human behavior model automatically and identify human’s behaviors without any manually annotated trainings.

  2. a Novel Deep Convolutional Neural Network for Spectral-Spatial Classification of Hyperspectral Data

    Science.gov (United States)

    Li, N.; Wang, C.; Zhao, H.; Gong, X.; Wang, D.

    2018-04-01

    Spatial and spectral information are obtained simultaneously by hyperspectral remote sensing. Joint extraction of these information of hyperspectral image is one of most import methods for hyperspectral image classification. In this paper, a novel deep convolutional neural network (CNN) is proposed, which extracts spectral-spatial information of hyperspectral images correctly. The proposed model not only learns sufficient knowledge from the limited number of samples, but also has powerful generalization ability. The proposed framework based on three-dimensional convolution can extract spectral-spatial features of labeled samples effectively. Though CNN has shown its robustness to distortion, it cannot extract features of different scales through the traditional pooling layer that only have one size of pooling window. Hence, spatial pyramid pooling (SPP) is introduced into three-dimensional local convolutional filters for hyperspectral classification. Experimental results with a widely used hyperspectral remote sensing dataset show that the proposed model provides competitive performance.

  3. Convolutional Dictionary Learning: Acceleration and Convergence

    Science.gov (United States)

    Chun, Il Yong; Fessler, Jeffrey A.

    2018-04-01

    Convolutional dictionary learning (CDL or sparsifying CDL) has many applications in image processing and computer vision. There has been growing interest in developing efficient algorithms for CDL, mostly relying on the augmented Lagrangian (AL) method or the variant alternating direction method of multipliers (ADMM). When their parameters are properly tuned, AL methods have shown fast convergence in CDL. However, the parameter tuning process is not trivial due to its data dependence and, in practice, the convergence of AL methods depends on the AL parameters for nonconvex CDL problems. To moderate these problems, this paper proposes a new practically feasible and convergent Block Proximal Gradient method using a Majorizer (BPG-M) for CDL. The BPG-M-based CDL is investigated with different block updating schemes and majorization matrix designs, and further accelerated by incorporating some momentum coefficient formulas and restarting techniques. All of the methods investigated incorporate a boundary artifacts removal (or, more generally, sampling) operator in the learning model. Numerical experiments show that, without needing any parameter tuning process, the proposed BPG-M approach converges more stably to desirable solutions of lower objective values than the existing state-of-the-art ADMM algorithm and its memory-efficient variant do. Compared to the ADMM approaches, the BPG-M method using a multi-block updating scheme is particularly useful in single-threaded CDL algorithm handling large datasets, due to its lower memory requirement and no polynomial computational complexity. Image denoising experiments show that, for relatively strong additive white Gaussian noise, the filters learned by BPG-M-based CDL outperform those trained by the ADMM approach.

  4. Lidar Cloud Detection with Fully Convolutional Networks

    Science.gov (United States)

    Cromwell, E.; Flynn, D.

    2017-12-01

    The vertical distribution of clouds from active remote sensing instrumentation is a widely used data product from global atmospheric measuring sites. The presence of clouds can be expressed as a binary cloud mask and is a primary input for climate modeling efforts and cloud formation studies. Current cloud detection algorithms producing these masks do not accurately identify the cloud boundaries and tend to oversample or over-represent the cloud. This translates as uncertainty for assessing the radiative impact of clouds and tracking changes in cloud climatologies. The Atmospheric Radiation Measurement (ARM) program has over 20 years of micro-pulse lidar (MPL) and High Spectral Resolution Lidar (HSRL) instrument data and companion automated cloud mask product at the mid-latitude Southern Great Plains (SGP) and the polar North Slope of Alaska (NSA) atmospheric observatory. Using this data, we train a fully convolutional network (FCN) with semi-supervised learning to segment lidar imagery into geometric time-height cloud locations for the SGP site and MPL instrument. We then use transfer learning to train a FCN for (1) the MPL instrument at the NSA site and (2) for the HSRL. In our semi-supervised approach, we pre-train the classification layers of the FCN with weakly labeled lidar data. Then, we facilitate end-to-end unsupervised pre-training and transition to fully supervised learning with ground truth labeled data. Our goal is to improve the cloud mask accuracy and precision for the MPL instrument to 95% and 80%, respectively, compared to the current cloud mask algorithms of 89% and 50%. For the transfer learning based FCN for the HSRL instrument, our goal is to achieve a cloud mask accuracy of 90% and a precision of 80%.

  5. Convolutional neural network architectures for predicting DNA–protein binding

    Science.gov (United States)

    Zeng, Haoyang; Edwards, Matthew D.; Liu, Ge; Gifford, David K.

    2016-01-01

    Motivation: Convolutional neural networks (CNN) have outperformed conventional methods in modeling the sequence specificity of DNA–protein binding. Yet inappropriate CNN architectures can yield poorer performance than simpler models. Thus an in-depth understanding of how to match CNN architecture to a given task is needed to fully harness the power of CNNs for computational biology applications. Results: We present a systematic exploration of CNN architectures for predicting DNA sequence binding using a large compendium of transcription factor datasets. We identify the best-performing architectures by varying CNN width, depth and pooling designs. We find that adding convolutional kernels to a network is important for motif-based tasks. We show the benefits of CNNs in learning rich higher-order sequence features, such as secondary motifs and local sequence context, by comparing network performance on multiple modeling tasks ranging in difficulty. We also demonstrate how careful construction of sequence benchmark datasets, using approaches that control potentially confounding effects like positional or motif strength bias, is critical in making fair comparisons between competing methods. We explore how to establish the sufficiency of training data for these learning tasks, and we have created a flexible cloud-based framework that permits the rapid exploration of alternative neural network architectures for problems in computational biology. Availability and Implementation: All the models analyzed are available at http://cnn.csail.mit.edu. Contact: gifford@mit.edu Supplementary information: Supplementary data are available at Bioinformatics online. PMID:27307608

  6. Texture synthesis using convolutional neural networks with long-range consistency and spectral constraints

    NARCIS (Netherlands)

    Schreiber, Shaun; Geldenhuys, Jaco; Villiers, De Hendrik

    2017-01-01

    Procedural texture generation enables the creation of more rich and detailed virtual environments without the help of an artist. However, finding a flexible generative model of real world textures remains an open problem. We present a novel Convolutional Neural Network based texture model

  7. Classifying images using restricted Boltzmann machines and convolutional neural networks

    Science.gov (United States)

    Zhao, Zhijun; Xu, Tongde; Dai, Chenyu

    2017-07-01

    To improve the feature recognition ability of deep model transfer learning, we propose a hybrid deep transfer learning method for image classification based on restricted Boltzmann machines (RBM) and convolutional neural networks (CNNs). It integrates learning abilities of two models, which conducts subject classification by exacting structural higher-order statistics features of images. While the method transfers the trained convolutional neural networks to the target datasets, fully-connected layers can be replaced by restricted Boltzmann machine layers; then the restricted Boltzmann machine layers and Softmax classifier are retrained, and BP neural network can be used to fine-tuned the hybrid model. The restricted Boltzmann machine layers has not only fully integrated the whole feature maps, but also learns the statistical features of target datasets in the view of the biggest logarithmic likelihood, thus removing the effects caused by the content differences between datasets. The experimental results show that the proposed method has improved the accuracy of image classification, outperforming other methods on Pascal VOC2007 and Caltech101 datasets.

  8. No-reference image quality assessment based on statistics of convolution feature maps

    Science.gov (United States)

    Lv, Xiaoxin; Qin, Min; Chen, Xiaohui; Wei, Guo

    2018-04-01

    We propose a Convolutional Feature Maps (CFM) driven approach to accurately predict image quality. Our motivation bases on the finding that the Nature Scene Statistic (NSS) features on convolution feature maps are significantly sensitive to distortion degree of an image. In our method, a Convolutional Neural Network (CNN) is trained to obtain kernels for generating CFM. We design a forward NSS layer which performs on CFM to better extract NSS features. The quality aware features derived from the output of NSS layer is effective to describe the distortion type and degree an image suffered. Finally, a Support Vector Regression (SVR) is employed in our No-Reference Image Quality Assessment (NR-IQA) model to predict a subjective quality score of a distorted image. Experiments conducted on two public databases demonstrate the promising performance of the proposed method is competitive to state of the art NR-IQA methods.

  9. Multi-Input Convolutional Neural Network for Flower Grading

    Directory of Open Access Journals (Sweden)

    Yu Sun

    2017-01-01

    Full Text Available Flower grading is a significant task because it is extremely convenient for managing the flowers in greenhouse and market. With the development of computer vision, flower grading has become an interdisciplinary focus in both botany and computer vision. A new dataset named BjfuGloxinia contains three quality grades; each grade consists of 107 samples and 321 images. A multi-input convolutional neural network is designed for large scale flower grading. Multi-input CNN achieves a satisfactory accuracy of 89.6% on the BjfuGloxinia after data augmentation. Compared with a single-input CNN, the accuracy of multi-input CNN is increased by 5% on average, demonstrating that multi-input convolutional neural network is a promising model for flower grading. Although data augmentation contributes to the model, the accuracy is still limited by lack of samples diversity. Majority of misclassification is derived from the medium class. The image processing based bud detection is useful for reducing the misclassification, increasing the accuracy of flower grading to approximately 93.9%.

  10. Siamese convolutional networks for tracking the spine motion

    Science.gov (United States)

    Liu, Yuan; Sui, Xiubao; Sun, Yicheng; Liu, Chengwei; Hu, Yong

    2017-09-01

    Deep learning models have demonstrated great success in various computer vision tasks such as image classification and object tracking. However, tracking the lumbar spine by digitalized video fluoroscopic imaging (DVFI), which can quantitatively analyze the motion mode of spine to diagnose lumbar instability, has not yet been well developed due to the lack of steady and robust tracking method. In this paper, we propose a novel visual tracking algorithm of the lumbar vertebra motion based on a Siamese convolutional neural network (CNN) model. We train a full-convolutional neural network offline to learn generic image features. The network is trained to learn a similarity function that compares the labeled target in the first frame with the candidate patches in the current frame. The similarity function returns a high score if the two images depict the same object. Once learned, the similarity function is used to track a previously unseen object without any adapting online. In the current frame, our tracker is performed by evaluating the candidate rotated patches sampled around the previous frame target position and presents a rotated bounding box to locate the predicted target precisely. Results indicate that the proposed tracking method can detect the lumbar vertebra steadily and robustly. Especially for images with low contrast and cluttered background, the presented tracker can still achieve good tracking performance. Further, the proposed algorithm operates at high speed for real time tracking.

  11. Convolutional neural networks for vibrational spectroscopic data analysis.

    Science.gov (United States)

    Acquarelli, Jacopo; van Laarhoven, Twan; Gerretzen, Jan; Tran, Thanh N; Buydens, Lutgarde M C; Marchiori, Elena

    2017-02-15

    In this work we show that convolutional neural networks (CNNs) can be efficiently used to classify vibrational spectroscopic data and identify important spectral regions. CNNs are the current state-of-the-art in image classification and speech recognition and can learn interpretable representations of the data. These characteristics make CNNs a good candidate for reducing the need for preprocessing and for highlighting important spectral regions, both of which are crucial steps in the analysis of vibrational spectroscopic data. Chemometric analysis of vibrational spectroscopic data often relies on preprocessing methods involving baseline correction, scatter correction and noise removal, which are applied to the spectra prior to model building. Preprocessing is a critical step because even in simple problems using 'reasonable' preprocessing methods may decrease the performance of the final model. We develop a new CNN based method and provide an accompanying publicly available software. It is based on a simple CNN architecture with a single convolutional layer (a so-called shallow CNN). Our method outperforms standard classification algorithms used in chemometrics (e.g. PLS) in terms of accuracy when applied to non-preprocessed test data (86% average accuracy compared to the 62% achieved by PLS), and it achieves better performance even on preprocessed test data (96% average accuracy compared to the 89% achieved by PLS). For interpretability purposes, our method includes a procedure for finding important spectral regions, thereby facilitating qualitative interpretation of results. Copyright © 2016 Elsevier B.V. All rights reserved.

  12. A convolutional neural network neutrino event classifier

    International Nuclear Information System (INIS)

    Aurisano, A.; Sousa, A.; Radovic, A.; Vahle, P.; Rocco, D.; Pawloski, G.; Himmel, A.; Niner, E.; Messier, M.D.; Psihas, F.

    2016-01-01

    Convolutional neural networks (CNNs) have been widely applied in the computer vision community to solve complex problems in image recognition and analysis. We describe an application of the CNN technology to the problem of identifying particle interactions in sampling calorimeters used commonly in high energy physics and high energy neutrino physics in particular. Following a discussion of the core concepts of CNNs and recent innovations in CNN architectures related to the field of deep learning, we outline a specific application to the NOvA neutrino detector. This algorithm, CVN (Convolutional Visual Network) identifies neutrino interactions based on their topology without the need for detailed reconstruction and outperforms algorithms currently in use by the NOvA collaboration.

  13. Quasi-cyclic unit memory convolutional codes

    DEFF Research Database (Denmark)

    Justesen, Jørn; Paaske, Erik; Ballan, Mark

    1990-01-01

    Unit memory convolutional codes with generator matrices, which are composed of circulant submatrices, are introduced. This structure facilitates the analysis of efficient search for good codes. Equivalences among such codes and some of the basic structural properties are discussed. In particular......, catastrophic encoders and minimal encoders are characterized and dual codes treated. Further, various distance measures are discussed, and a number of good codes, some of which result from efficient computer search and some of which result from known block codes, are presented...

  14. Applying Gradient Descent in Convolutional Neural Networks

    Science.gov (United States)

    Cui, Nan

    2018-04-01

    With the development of the integrated circuit and computer science, people become caring more about solving practical issues via information technologies. Along with that, a new subject called Artificial Intelligent (AI) comes up. One popular research interest of AI is about recognition algorithm. In this paper, one of the most common algorithms, Convolutional Neural Networks (CNNs) will be introduced, for image recognition. Understanding its theory and structure is of great significance for every scholar who is interested in this field. Convolution Neural Network is an artificial neural network which combines the mathematical method of convolution and neural network. The hieratical structure of CNN provides it reliable computer speed and reasonable error rate. The most significant characteristics of CNNs are feature extraction, weight sharing and dimension reduction. Meanwhile, combining with the Back Propagation (BP) mechanism and the Gradient Descent (GD) method, CNNs has the ability to self-study and in-depth learning. Basically, BP provides an opportunity for backwardfeedback for enhancing reliability and GD is used for self-training process. This paper mainly discusses the CNN and the related BP and GD algorithms, including the basic structure and function of CNN, details of each layer, the principles and features of BP and GD, and some examples in practice with a summary in the end.

  15. Phylogenetic convolutional neural networks in metagenomics.

    Science.gov (United States)

    Fioravanti, Diego; Giarratano, Ylenia; Maggio, Valerio; Agostinelli, Claudio; Chierici, Marco; Jurman, Giuseppe; Furlanello, Cesare

    2018-03-08

    Convolutional Neural Networks can be effectively used only when data are endowed with an intrinsic concept of neighbourhood in the input space, as is the case of pixels in images. We introduce here Ph-CNN, a novel deep learning architecture for the classification of metagenomics data based on the Convolutional Neural Networks, with the patristic distance defined on the phylogenetic tree being used as the proximity measure. The patristic distance between variables is used together with a sparsified version of MultiDimensional Scaling to embed the phylogenetic tree in a Euclidean space. Ph-CNN is tested with a domain adaptation approach on synthetic data and on a metagenomics collection of gut microbiota of 38 healthy subjects and 222 Inflammatory Bowel Disease patients, divided in 6 subclasses. Classification performance is promising when compared to classical algorithms like Support Vector Machines and Random Forest and a baseline fully connected neural network, e.g. the Multi-Layer Perceptron. Ph-CNN represents a novel deep learning approach for the classification of metagenomics data. Operatively, the algorithm has been implemented as a custom Keras layer taking care of passing to the following convolutional layer not only the data but also the ranked list of neighbourhood of each sample, thus mimicking the case of image data, transparently to the user.

  16. Image quality assessment using deep convolutional networks

    Science.gov (United States)

    Li, Yezhou; Ye, Xiang; Li, Yong

    2017-12-01

    This paper proposes a method of accurately assessing image quality without a reference image by using a deep convolutional neural network. Existing training based methods usually utilize a compact set of linear filters for learning features of images captured by different sensors to assess their quality. These methods may not be able to learn the semantic features that are intimately related with the features used in human subject assessment. Observing this drawback, this work proposes training a deep convolutional neural network (CNN) with labelled images for image quality assessment. The ReLU in the CNN allows non-linear transformations for extracting high-level image features, providing a more reliable assessment of image quality than linear filters. To enable the neural network to take images of any arbitrary size as input, the spatial pyramid pooling (SPP) is introduced connecting the top convolutional layer and the fully-connected layer. In addition, the SPP makes the CNN robust to object deformations to a certain extent. The proposed method taking an image as input carries out an end-to-end learning process, and outputs the quality of the image. It is tested on public datasets. Experimental results show that it outperforms existing methods by a large margin and can accurately assess the image quality on images taken by different sensors of varying sizes.

  17. Paediatric frontal chest radiograph screening with fine-tuned convolutional neural networks

    CSIR Research Space (South Africa)

    Gerrand, Jonathan D

    2017-07-01

    Full Text Available of fine-tuned convolutional neural networks (CNN). We use two popular CNN models that are pre-trained on a large natural image dataset and two distinct datasets containing paediatric and adult radiographs respectively. Evaluation is performed using a 5...

  18. Convolutions of Heavy Tailed Random Variables and Applications to Portfolio Diversification and MA(1) Time Series

    NARCIS (Netherlands)

    J.L. Geluk (Jaap); L. Peng (Liang); C.G. de Vries (Casper)

    1999-01-01

    textabstractThe paper characterizes first and second order tail behavior of convolutions of i.i.d. heavy tailed random variables with support on the real line. The result is applied to the problem of risk diversification in portfolio analysis and to the estimation of the parameter in a MA(1) model.

  19. Multi-Branch Fully Convolutional Network for Face Detection

    KAUST Repository

    Bai, Yancheng

    2017-07-20

    Face detection is a fundamental problem in computer vision. It is still a challenging task in unconstrained conditions due to significant variations in scale, pose, expressions, and occlusion. In this paper, we propose a multi-branch fully convolutional network (MB-FCN) for face detection, which considers both efficiency and effectiveness in the design process. Our MB-FCN detector can deal with faces at all scale ranges with only a single pass through the backbone network. As such, our MB-FCN model saves computation and thus is more efficient, compared to previous methods that make multiple passes. For each branch, the specific skip connections of the convolutional feature maps at different layers are exploited to represent faces in specific scale ranges. Specifically, small faces can be represented with both shallow fine-grained and deep powerful coarse features. With this representation, superior improvement in performance is registered for the task of detecting small faces. We test our MB-FCN detector on two public face detection benchmarks, including FDDB and WIDER FACE. Extensive experiments show that our detector outperforms state-of-the-art methods on all these datasets in general and by a substantial margin on the most challenging among them (e.g. WIDER FACE Hard subset). Also, MB-FCN runs at 15 FPS on a GPU for images of size 640 x 480 with no assumption on the minimum detectable face size.

  20. Transforming Musical Signals through a Genre Classifying Convolutional Neural Network

    Science.gov (United States)

    Geng, S.; Ren, G.; Ogihara, M.

    2017-05-01

    Convolutional neural networks (CNNs) have been successfully applied on both discriminative and generative modeling for music-related tasks. For a particular task, the trained CNN contains information representing the decision making or the abstracting process. One can hope to manipulate existing music based on this 'informed' network and create music with new features corresponding to the knowledge obtained by the network. In this paper, we propose a method to utilize the stored information from a CNN trained on musical genre classification task. The network was composed of three convolutional layers, and was trained to classify five-second song clips into five different genres. After training, randomly selected clips were modified by maximizing the sum of outputs from the network layers. In addition to the potential of such CNNs to produce interesting audio transformation, more information about the network and the original music could be obtained from the analysis of the generated features since these features indicate how the network 'understands' the music.

  1. Classification of decays involving variable decay chains with convolutional architectures

    CERN Multimedia

    CERN. Geneva

    2018-01-01

    Vidyo contribution We present a technique to perform classification of decays that exhibit decay chains involving a variable number of particles, which include a broad class of $B$ meson decays sensitive to new physics. The utility of such decays as a probe of the Standard Model is dependent upon accurate determination of the decay rate, which is challenged by the combinatorial background arising in high-multiplicity decay modes. In our model, each particle in the decay event is represented as a fixed-dimensional vector of feature attributes, forming an $n \\times k$ representation of the event, where $n$ is the number of particles in the event and $k$ is the dimensionality of the feature vector. A convolutional architecture is used to capture dependencies between the embedded particle representations and perform the final classification. The proposed model performs outperforms standard machine learning approaches based on Monte Carlo studies across a range of variable final-state decays with the Belle II det...

  2. Network Intrusion Detection through Stacking Dilated Convolutional Autoencoders

    Directory of Open Access Journals (Sweden)

    Yang Yu

    2017-01-01

    Full Text Available Network intrusion detection is one of the most important parts for cyber security to protect computer systems against malicious attacks. With the emergence of numerous sophisticated and new attacks, however, network intrusion detection techniques are facing several significant challenges. The overall objective of this study is to learn useful feature representations automatically and efficiently from large amounts of unlabeled raw network traffic data by using deep learning approaches. We propose a novel network intrusion model by stacking dilated convolutional autoencoders and evaluate our method on two new intrusion detection datasets. Several experiments were carried out to check the effectiveness of our approach. The comparative experimental results demonstrate that the proposed model can achieve considerably high performance which meets the demand of high accuracy and adaptability of network intrusion detection systems (NIDSs. It is quite potential and promising to apply our model in the large-scale and real-world network environments.

  3. Static facial expression recognition with convolution neural networks

    Science.gov (United States)

    Zhang, Feng; Chen, Zhong; Ouyang, Chao; Zhang, Yifei

    2018-03-01

    Facial expression recognition is a currently active research topic in the fields of computer vision, pattern recognition and artificial intelligence. In this paper, we have developed a convolutional neural networks (CNN) for classifying human emotions from static facial expression into one of the seven facial emotion categories. We pre-train our CNN model on the combined FER2013 dataset formed by train, validation and test set and fine-tune on the extended Cohn-Kanade database. In order to reduce the overfitting of the models, we utilized different techniques including dropout and batch normalization in addition to data augmentation. According to the experimental result, our CNN model has excellent classification performance and robustness for facial expression recognition.

  4. An Algorithm for the Convolution of Legendre Series

    KAUST Repository

    Hale, Nicholas; Townsend, Alex

    2014-01-01

    An O(N2) algorithm for the convolution of compactly supported Legendre series is described. The algorithm is derived from the convolution theorem for Legendre polynomials and the recurrence relation satisfied by spherical Bessel functions. Combining with previous work yields an O(N 2) algorithm for the convolution of Chebyshev series. Numerical results are presented to demonstrate the improved efficiency over the existing algorithm. © 2014 Society for Industrial and Applied Mathematics.

  5. The Urbanik generalized convolutions in the non-commutative ...

    Indian Academy of Sciences (India)

    −sν(dx) < ∞. Now we apply this construction to the Kendall convolution case, starting with the weakly stable measure δ1. Example 1. Let △ be the Kendall convolution, i.e. the generalized convolution with the probability kernel: δ1△δa = (1 − a)δ1 + aπ2 for a ∈ [0, 1] and π2 be the Pareto distribution with the density π2(dx) =.

  6. On a Generalized Hankel Type Convolution of Generalized Functions

    Indian Academy of Sciences (India)

    Generalized Hankel type transformation; Parserval relation; generalized ... The classical generalized Hankel type convolution are defined and extended to a class of generalized functions. ... Proceedings – Mathematical Sciences | News.

  7. Enhanced online convolutional neural networks for object tracking

    Science.gov (United States)

    Zhang, Dengzhuo; Gao, Yun; Zhou, Hao; Li, Tianwen

    2018-04-01

    In recent several years, object tracking based on convolution neural network has gained more and more attention. The initialization and update of convolution filters can directly affect the precision of object tracking effective. In this paper, a novel object tracking via an enhanced online convolution neural network without offline training is proposed, which initializes the convolution filters by a k-means++ algorithm and updates the filters by an error back-propagation. The comparative experiments of 7 trackers on 15 challenging sequences showed that our tracker can perform better than other trackers in terms of AUC and precision.

  8. Cascaded K-means convolutional feature learner and its application to face recognition

    Science.gov (United States)

    Zhou, Daoxiang; Yang, Dan; Zhang, Xiaohong; Huang, Sheng; Feng, Shu

    2017-09-01

    Currently, considerable efforts have been devoted to devise image representation. However, handcrafted methods need strong domain knowledge and show low generalization ability, and conventional feature learning methods require enormous training data and rich parameters tuning experience. A lightened feature learner is presented to solve these problems with application to face recognition, which shares similar topology architecture as a convolutional neural network. Our model is divided into three components: cascaded convolution filters bank learning layer, nonlinear processing layer, and feature pooling layer. Specifically, in the filters learning layer, we use K-means to learn convolution filters. Features are extracted via convoluting images with the learned filters. Afterward, in the nonlinear processing layer, hyperbolic tangent is employed to capture the nonlinear feature. In the feature pooling layer, to remove the redundancy information and incorporate the spatial layout, we exploit multilevel spatial pyramid second-order pooling technique to pool the features in subregions and concatenate them together as the final representation. Extensive experiments on four representative datasets demonstrate the effectiveness and robustness of our model to various variations, yielding competitive recognition results on extended Yale B and FERET. In addition, our method achieves the best identification performance on AR and labeled faces in the wild datasets among the comparative methods.

  9. Finger vein recognition based on convolutional neural network

    Directory of Open Access Journals (Sweden)

    Meng Gesi

    2017-01-01

    Full Text Available Biometric Authentication Technology has been widely used in this information age. As one of the most important technology of authentication, finger vein recognition attracts our attention because of its high security, reliable accuracy and excellent performance. However, the current finger vein recognition system is difficult to be applied widely because its complicated image pre-processing and not representative feature vectors. To solve this problem, a finger vein recognition method based on the convolution neural network (CNN is proposed in the paper. The image samples are directly input into the CNN model to extract its feature vector so that we can make authentication by comparing the Euclidean distance between these vectors. Finally, the Deep Learning Framework Caffe is adopted to verify this method. The result shows that there are great improvements in both speed and accuracy rate compared to the previous research. And the model has nice robustness in illumination and rotation.

  10. Convolutive Blind Source Separation Methods

    DEFF Research Database (Denmark)

    Pedersen, Michael Syskind; Larsen, Jan; Kjems, Ulrik

    2008-01-01

    During the past decades, much attention has been given to the separation of mixed sources, in particular for the blind case where both the sources and the mixing process are unknown and only recordings of the mixtures are available. In several situations it is desirable to recover all sources from...... the recorded mixtures, or at least to segregate a particular source. Furthermore, it may be useful to identify the mixing process itself to reveal information about the physical mixing system. In some simple mixing models each recording consists of a sum of differently weighted source signals. However, in many...... real-world applications, such as in acoustics, the mixing process is more complex. In such systems, the mixtures are weighted and delayed, and each source contributes to the sum with multiple delays corresponding to the multiple paths by which an acoustic signal propagates to a microphone...

  11. Convolutional neural networks and face recognition task

    Science.gov (United States)

    Sochenkova, A.; Sochenkov, I.; Makovetskii, A.; Vokhmintsev, A.; Melnikov, A.

    2017-09-01

    Computer vision tasks are remaining very important for the last couple of years. One of the most complicated problems in computer vision is face recognition that could be used in security systems to provide safety and to identify person among the others. There is a variety of different approaches to solve this task, but there is still no universal solution that would give adequate results in some cases. Current paper presents following approach. Firstly, we extract an area containing face, then we use Canny edge detector. On the next stage we use convolutional neural networks (CNN) to finally solve face recognition and person identification task.

  12. Fourier transforms and convolutions for the experimentalist

    CERN Document Server

    Jennison, RC

    1961-01-01

    Fourier Transforms and Convolutions for the Experimentalist provides the experimentalist with a guide to the principles and practical uses of the Fourier transformation. It aims to bridge the gap between the more abstract account of a purely mathematical approach and the rule of thumb calculation and intuition of the practical worker. The monograph springs from a lecture course which the author has given in recent years and for which he has drawn upon a number of sources, including a set of notes compiled by the late Dr. I. C. Browne from a series of lectures given by Mr. J . A. Ratcliffe of t

  13. Target recognition based on convolutional neural network

    Science.gov (United States)

    Wang, Liqiang; Wang, Xin; Xi, Fubiao; Dong, Jian

    2017-11-01

    One of the important part of object target recognition is the feature extraction, which can be classified into feature extraction and automatic feature extraction. The traditional neural network is one of the automatic feature extraction methods, while it causes high possibility of over-fitting due to the global connection. The deep learning algorithm used in this paper is a hierarchical automatic feature extraction method, trained with the layer-by-layer convolutional neural network (CNN), which can extract the features from lower layers to higher layers. The features are more discriminative and it is beneficial to the object target recognition.

  14. QCDNUM: Fast QCD evolution and convolution

    Science.gov (United States)

    Botje, M.

    2011-02-01

    The QCDNUM program numerically solves the evolution equations for parton densities and fragmentation functions in perturbative QCD. Un-polarised parton densities can be evolved up to next-to-next-to-leading order in powers of the strong coupling constant, while polarised densities or fragmentation functions can be evolved up to next-to-leading order. Other types of evolution can be accessed by feeding alternative sets of evolution kernels into the program. A versatile convolution engine provides tools to compute parton luminosities, cross-sections in hadron-hadron scattering, and deep inelastic structure functions in the zero-mass scheme or in generalised mass schemes. Input to these calculations are either the QCDNUM evolved densities, or those read in from an external parton density repository. Included in the software distribution are packages to calculate zero-mass structure functions in un-polarised deep inelastic scattering, and heavy flavour contributions to these structure functions in the fixed flavour number scheme. Program summaryProgram title: QCDNUM version: 17.00 Catalogue identifier: AEHV_v1_0 Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AEHV_v1_0.html Program obtainable from: CPC Program Library, Queen's University, Belfast, N. Ireland Licensing provisions: GNU Public Licence No. of lines in distributed program, including test data, etc.: 45 736 No. of bytes in distributed program, including test data, etc.: 911 569 Distribution format: tar.gz Programming language: Fortran-77 Computer: All Operating system: All RAM: Typically 3 Mbytes Classification: 11.5 Nature of problem: Evolution of the strong coupling constant and parton densities, up to next-to-next-to-leading order in perturbative QCD. Computation of observable quantities by Mellin convolution of the evolved densities with partonic cross-sections. Solution method: Parametrisation of the parton densities as linear or quadratic splines on a discrete grid, and evolution of the spline

  15. Improving the Separability of Deep Features with Discriminative Convolution Filters for RSI Classification

    Directory of Open Access Journals (Sweden)

    Na Liu

    2018-03-01

    Full Text Available The extraction of activation vectors (or deep features from the fully connected layers of a convolutional neural network (CNN model is widely used for remote sensing image (RSI representation. In this study, we propose to learn discriminative convolution filter (DCF based on class-specific separability criteria for linear transformation of deep features. In particular, two types of pretrained CNN called CaffeNet and VGG-VD16 are introduced to illustrate the generality of the proposed DCF. The activation vectors extracted from the fully connected layers of a CNN are rearranged into the form of an image matrix, from which a spatial arrangement of local patches is extracted using sliding window strategy. DCF learning is then performed on each local patch individually to obtain the corresponding discriminative convolution kernel through generalized eigenvalue decomposition. The proposed DCF learning characterizes that a convolutional kernel with small size (e.g., 3 × 3 pixels can be effectively learned on a small-size local patch (e.g., 8 × 8 pixels, thereby ensuring that the linear transformation of deep features can maintain low computational complexity. Experiments on two RSI datasets demonstrate the effectiveness of DCF in improving the classification performances of deep features without increasing dimensionality.

  16. Spectral-spatial classification of hyperspectral image using three-dimensional convolution network

    Science.gov (United States)

    Liu, Bing; Yu, Xuchu; Zhang, Pengqiang; Tan, Xiong; Wang, Ruirui; Zhi, Lu

    2018-01-01

    Recently, hyperspectral image (HSI) classification has become a focus of research. However, the complex structure of an HSI makes feature extraction difficult to achieve. Most current methods build classifiers based on complex handcrafted features computed from the raw inputs. The design of an improved 3-D convolutional neural network (3D-CNN) model for HSI classification is described. This model extracts features from both the spectral and spatial dimensions through the application of 3-D convolutions, thereby capturing the important discrimination information encoded in multiple adjacent bands. The designed model views the HSI cube data altogether without relying on any pre- or postprocessing. In addition, the model is trained in an end-to-end fashion without any handcrafted features. The designed model was applied to three widely used HSI datasets. The experimental results demonstrate that the 3D-CNN-based method outperforms conventional methods even with limited labeled training samples.

  17. Modified Stieltjes Transform and Generalized Convolutions of Probability Distributions

    Directory of Open Access Journals (Sweden)

    Lev B. Klebanov

    2018-01-01

    Full Text Available The classical Stieltjes transform is modified in such a way as to generalize both Stieltjes and Fourier transforms. This transform allows the introduction of new classes of commutative and non-commutative generalized convolutions. A particular case of such a convolution for degenerate distributions appears to be the Wigner semicircle distribution.

  18. Color encoding in biologically-inspired convolutional neural networks.

    Science.gov (United States)

    Rafegas, Ivet; Vanrell, Maria

    2018-05-11

    Convolutional Neural Networks have been proposed as suitable frameworks to model biological vision. Some of these artificial networks showed representational properties that rival primate performances in object recognition. In this paper we explore how color is encoded in a trained artificial network. It is performed by estimating a color selectivity index for each neuron, which allows us to describe the neuron activity to a color input stimuli. The index allows us to classify whether they are color selective or not and if they are of a single or double color. We have determined that all five convolutional layers of the network have a large number of color selective neurons. Color opponency clearly emerges in the first layer, presenting 4 main axes (Black-White, Red-Cyan, Blue-Yellow and Magenta-Green), but this is reduced and rotated as we go deeper into the network. In layer 2 we find a denser hue sampling of color neurons and opponency is reduced almost to one new main axis, the Bluish-Orangish coinciding with the dataset bias. In layers 3, 4 and 5 color neurons are similar amongst themselves, presenting different type of neurons that detect specific colored objects (e.g., orangish faces), specific surrounds (e.g., blue sky) or specific colored or contrasted object-surround configurations (e.g. blue blob in a green surround). Overall, our work concludes that color and shape representation are successively entangled through all the layers of the studied network, revealing certain parallelisms with the reported evidences in primate brains that can provide useful insight into intermediate hierarchical spatio-chromatic representations. Copyright © 2018 Elsevier Ltd. All rights reserved.

  19. Fully convolutional neural networks improve abdominal organ segmentation

    Science.gov (United States)

    Bobo, Meg F.; Bao, Shunxing; Huo, Yuankai; Yao, Yuang; Virostko, Jack; Plassard, Andrew J.; Lyu, Ilwoo; Assad, Albert; Abramson, Richard G.; Hilmes, Melissa A.; Landman, Bennett A.

    2018-03-01

    Abdominal image segmentation is a challenging, yet important clinical problem. Variations in body size, position, and relative organ positions greatly complicate the segmentation process. Historically, multi-atlas methods have achieved leading results across imaging modalities and anatomical targets. However, deep learning is rapidly overtaking classical approaches for image segmentation. Recently, Zhou et al. showed that fully convolutional networks produce excellent results in abdominal organ segmentation of computed tomography (CT) scans. Yet, deep learning approaches have not been applied to whole abdomen magnetic resonance imaging (MRI) segmentation. Herein, we evaluate the applicability of an existing fully convolutional neural network (FCNN) designed for CT imaging to segment abdominal organs on T2 weighted (T2w) MRI's with two examples. In the primary example, we compare a classical multi-atlas approach with FCNN on forty-five T2w MRI's acquired from splenomegaly patients with five organs labeled (liver, spleen, left kidney, right kidney, and stomach). Thirty-six images were used for training while nine were used for testing. The FCNN resulted in a Dice similarity coefficient (DSC) of 0.930 in spleens, 0.730 in left kidneys, 0.780 in right kidneys, 0.913 in livers, and 0.556 in stomachs. The performance measures for livers, spleens, right kidneys, and stomachs were significantly better than multi-atlas (p < 0.05, Wilcoxon rank-sum test). In a secondary example, we compare the multi-atlas approach with FCNN on 138 distinct T2w MRI's with manually labeled pancreases (one label). On the pancreas dataset, the FCNN resulted in a median DSC of 0.691 in pancreases versus 0.287 for multi-atlas. The results are highly promising given relatively limited training data and without specific training of the FCNN model and illustrate the potential of deep learning approaches to transcend imaging modalities. 1

  20. Hierarchical graphical-based human pose estimation via local multi-resolution convolutional neural network

    Science.gov (United States)

    Zhu, Aichun; Wang, Tian; Snoussi, Hichem

    2018-03-01

    This paper addresses the problems of the graphical-based human pose estimation in still images, including the diversity of appearances and confounding background clutter. We present a new architecture for estimating human pose using a Convolutional Neural Network (CNN). Firstly, a Relative Mixture Deformable Model (RMDM) is defined by each pair of connected parts to compute the relative spatial information in the graphical model. Secondly, a Local Multi-Resolution Convolutional Neural Network (LMR-CNN) is proposed to train and learn the multi-scale representation of each body parts by combining different levels of part context. Thirdly, a LMR-CNN based hierarchical model is defined to explore the context information of limb parts. Finally, the experimental results demonstrate the effectiveness of the proposed deep learning approach for human pose estimation.

  1. Hierarchical graphical-based human pose estimation via local multi-resolution convolutional neural network

    Directory of Open Access Journals (Sweden)

    Aichun Zhu

    2018-03-01

    Full Text Available This paper addresses the problems of the graphical-based human pose estimation in still images, including the diversity of appearances and confounding background clutter. We present a new architecture for estimating human pose using a Convolutional Neural Network (CNN. Firstly, a Relative Mixture Deformable Model (RMDM is defined by each pair of connected parts to compute the relative spatial information in the graphical model. Secondly, a Local Multi-Resolution Convolutional Neural Network (LMR-CNN is proposed to train and learn the multi-scale representation of each body parts by combining different levels of part context. Thirdly, a LMR-CNN based hierarchical model is defined to explore the context information of limb parts. Finally, the experimental results demonstrate the effectiveness of the proposed deep learning approach for human pose estimation.

  2. The quick convolution of galaxy profiles, with application to power-law intensity distributions

    International Nuclear Information System (INIS)

    Bailey, M.E.; Sparks, W.B.

    1983-01-01

    The two-dimensional convolution of a circularly symmetric galaxy model with a Gaussian point-spread function of dispersion σ reduces to a single integral. This is solved analytically for models with power-law intensity distributions and results are given which relate the apparent core radius to σ and the power-law index k. The convolution integral is also simplified for the case of a point-spread function corresponding to a circular aperture. Models of galactic nuclei with stellar density cusps can only be distinguished from alternatives with small core radii if both the brightness and seeing profiles are measured accurately. The results are applied to data on the light distribution at the Galactic Centre. (author)

  3. Histopathological Breast-Image Classification Using Local and Frequency Domains by Convolutional Neural Network

    Directory of Open Access Journals (Sweden)

    Abdullah-Al Nahid

    2018-01-01

    Full Text Available Identification of the malignancy of tissues from Histopathological images has always been an issue of concern to doctors and radiologists. This task is time-consuming, tedious and moreover very challenging. Success in finding malignancy from Histopathological images primarily depends on long-term experience, though sometimes experts disagree on their decisions. However, Computer Aided Diagnosis (CAD techniques help the radiologist to give a second opinion that can increase the reliability of the radiologist’s decision. Among the different image analysis techniques, classification of the images has always been a challenging task. Due to the intense complexity of biomedical images, it is always very challenging to provide a reliable decision about an image. The state-of-the-art Convolutional Neural Network (CNN technique has had great success in natural image classification. Utilizing advanced engineering techniques along with the CNN, in this paper, we have classified a set of Histopathological Breast-Cancer (BC images utilizing a state-of-the-art CNN model containing a residual block. Conventional CNN operation takes raw images as input and extracts the global features; however, the object oriented local features also contain significant information—for example, the Local Binary Pattern (LBP represents the effective textural information, Histogram represent the pixel strength distribution, Contourlet Transform (CT gives much detailed information about the smoothness about the edges, and Discrete Fourier Transform (DFT derives frequency-domain information from the image. Utilizing these advantages, along with our proposed novel CNN model, we have examined the performance of the novel CNN model as Histopathological image classifier. To do so, we have introduced five cases: (a Convolutional Neural Network Raw Image (CNN-I; (b Convolutional Neural Network CT Histogram (CNN-CH; (c Convolutional Neural Network CT LBP (CNN-CL; (d Convolutional

  4. Robust Vehicle Detection in Aerial Images Based on Cascaded Convolutional Neural Networks.

    Science.gov (United States)

    Zhong, Jiandan; Lei, Tao; Yao, Guangle

    2017-11-24

    Vehicle detection in aerial images is an important and challenging task. Traditionally, many target detection models based on sliding-window fashion were developed and achieved acceptable performance, but these models are time-consuming in the detection phase. Recently, with the great success of convolutional neural networks (CNNs) in computer vision, many state-of-the-art detectors have been designed based on deep CNNs. However, these CNN-based detectors are inefficient when applied in aerial image data due to the fact that the existing CNN-based models struggle with small-size object detection and precise localization. To improve the detection accuracy without decreasing speed, we propose a CNN-based detection model combining two independent convolutional neural networks, where the first network is applied to generate a set of vehicle-like regions from multi-feature maps of different hierarchies and scales. Because the multi-feature maps combine the advantage of the deep and shallow convolutional layer, the first network performs well on locating the small targets in aerial image data. Then, the generated candidate regions are fed into the second network for feature extraction and decision making. Comprehensive experiments are conducted on the Vehicle Detection in Aerial Imagery (VEDAI) dataset and Munich vehicle dataset. The proposed cascaded detection model yields high performance, not only in detection accuracy but also in detection speed.

  5. Deformable image registration using convolutional neural networks

    Science.gov (United States)

    Eppenhof, Koen A. J.; Lafarge, Maxime W.; Moeskops, Pim; Veta, Mitko; Pluim, Josien P. W.

    2018-03-01

    Deformable image registration can be time-consuming and often needs extensive parameterization to perform well on a specific application. We present a step towards a registration framework based on a three-dimensional convolutional neural network. The network directly learns transformations between pairs of three-dimensional images. The outputs of the network are three maps for the x, y, and z components of a thin plate spline transformation grid. The network is trained on synthetic random transformations, which are applied to a small set of representative images for the desired application. Training therefore does not require manually annotated ground truth deformation information. The methodology is demonstrated on public data sets of inspiration-expiration lung CT image pairs, which come with annotated corresponding landmarks for evaluation of the registration accuracy. Advantages of this methodology are its fast registration times and its minimal parameterization.

  6. Codeword Structure Analysis for LDPC Convolutional Codes

    Directory of Open Access Journals (Sweden)

    Hua Zhou

    2015-12-01

    Full Text Available The codewords of a low-density parity-check (LDPC convolutional code (LDPC-CC are characterised into structured and non-structured. The number of the structured codewords is dominated by the size of the polynomial syndrome former matrix H T ( D , while the number of the non-structured ones depends on the particular monomials or polynomials in H T ( D . By evaluating the relationship of the codewords between the mother code and its super codes, the low weight non-structured codewords in the super codes can be eliminated by appropriately choosing the monomials or polynomials in H T ( D , resulting in improved distance spectrum of the mother code.

  7. A Note on the Tail Behavior of Randomly Weighted Sums with Convolution-Equivalently Distributed Random Variables

    Directory of Open Access Journals (Sweden)

    Yang Yang

    2013-01-01

    Full Text Available We investigate the tailed asymptotic behavior of the randomly weighted sums with increments with convolution-equivalent distributions. Our obtained result can be directly applied to a discrete-time insurance risk model with insurance and financial risks and derive the asymptotics for the finite-time probability of the above risk model.

  8. An Improved Convolutional Neural Network on Crowd Density Estimation

    Directory of Open Access Journals (Sweden)

    Pan Shao-Yun

    2016-01-01

    Full Text Available In this paper, a new method is proposed for crowd density estimation. An improved convolutional neural network is combined with traditional texture feature. The data calculated by the convolutional layer can be treated as a new kind of features.So more useful information of images can be extracted by different features.In the meantime, the size of image has little effect on the result of convolutional neural network. Experimental results indicate that our scheme has adequate performance to allow for its use in real world applications.

  9. A Parallel Strategy for Convolutional Neural Network Based on Heterogeneous Cluster for Mobile Information System

    Directory of Open Access Journals (Sweden)

    Jilin Zhang

    2017-01-01

    Full Text Available With the development of the mobile systems, we gain a lot of benefits and convenience by leveraging mobile devices; at the same time, the information gathered by smartphones, such as location and environment, is also valuable for business to provide more intelligent services for customers. More and more machine learning methods have been used in the field of mobile information systems to study user behavior and classify usage patterns, especially convolutional neural network. With the increasing of model training parameters and data scale, the traditional single machine training method cannot meet the requirements of time complexity in practical application scenarios. The current training framework often uses simple data parallel or model parallel method to speed up the training process, which is why heterogeneous computing resources have not been fully utilized. To solve these problems, our paper proposes a delay synchronization convolutional neural network parallel strategy, which leverages the heterogeneous system. The strategy is based on both synchronous parallel and asynchronous parallel approaches; the model training process can reduce the dependence on the heterogeneous architecture in the premise of ensuring the model convergence, so the convolution neural network framework is more adaptive to different heterogeneous system environments. The experimental results show that the proposed delay synchronization strategy can achieve at least three times the speedup compared to the traditional data parallelism.

  10. DeepFix: A Fully Convolutional Neural Network for Predicting Human Eye Fixations.

    Science.gov (United States)

    Kruthiventi, Srinivas S S; Ayush, Kumar; Babu, R Venkatesh

    2017-09-01

    Understanding and predicting the human visual attention mechanism is an active area of research in the fields of neuroscience and computer vision. In this paper, we propose DeepFix, a fully convolutional neural network, which models the bottom-up mechanism of visual attention via saliency prediction. Unlike classical works, which characterize the saliency map using various hand-crafted features, our model automatically learns features in a hierarchical fashion and predicts the saliency map in an end-to-end manner. DeepFix is designed to capture semantics at multiple scales while taking global context into account, by using network layers with very large receptive fields. Generally, fully convolutional nets are spatially invariant-this prevents them from modeling location-dependent patterns (e.g., centre-bias). Our network handles this by incorporating a novel location-biased convolutional layer. We evaluate our model on multiple challenging saliency data sets and show that it achieves the state-of-the-art results.

  11. A Convolution-LSTM-Based Deep Neural Network for Cross-Domain MOOC Forum Post Classification

    Directory of Open Access Journals (Sweden)

    Xiaocong Wei

    2017-07-01

    Full Text Available Learners in a massive open online course often express feelings, exchange ideas and seek help by posting questions in discussion forums. Due to the very high learner-to-instructor ratios, it is unrealistic to expect instructors to adequately track the forums, find all of the issues that need resolution and understand their urgency and sentiment. In this paper, considering the biases among different courses, we propose a transfer learning framework based on a convolutional neural network and a long short-term memory model, called ConvL, to automatically identify whether a post expresses confusion, determine the urgency and classify the polarity of the sentiment. First, we learn the feature representation for each word by considering the local contextual feature via the convolution operation. Second, we learn the post representation from the features extracted through the convolution operation via the LSTM model, which considers the long-term temporal semantic relationships of features. Third, we investigate the possibility of transferring parameters from a model trained on one course to another course and the subsequent fine-tuning. Experiments on three real-world MOOC courses confirm the effectiveness of our framework. This work suggests that our model can potentially significantly increase the effectiveness of monitoring MOOC forums in real time.

  12. Village Building Identification Based on Ensemble Convolutional Neural Networks

    Science.gov (United States)

    Guo, Zhiling; Chen, Qi; Xu, Yongwei; Shibasaki, Ryosuke; Shao, Xiaowei

    2017-01-01

    In this study, we present the Ensemble Convolutional Neural Network (ECNN), an elaborate CNN frame formulated based on ensembling state-of-the-art CNN models, to identify village buildings from open high-resolution remote sensing (HRRS) images. First, to optimize and mine the capability of CNN for village mapping and to ensure compatibility with our classification targets, a few state-of-the-art models were carefully optimized and enhanced based on a series of rigorous analyses and evaluations. Second, rather than directly implementing building identification by using these models, we exploited most of their advantages by ensembling their feature extractor parts into a stronger model called ECNN based on the multiscale feature learning method. Finally, the generated ECNN was applied to a pixel-level classification frame to implement object identification. The proposed method can serve as a viable tool for village building identification with high accuracy and efficiency. The experimental results obtained from the test area in Savannakhet province, Laos, prove that the proposed ECNN model significantly outperforms existing methods, improving overall accuracy from 96.64% to 99.26%, and kappa from 0.57 to 0.86. PMID:29084154

  13. Transfer Learning for Video Recognition with Scarce Training Data for Deep Convolutional Neural Network

    OpenAIRE

    Su, Yu-Chuan; Chiu, Tzu-Hsuan; Yeh, Chun-Yen; Huang, Hsin-Fu; Hsu, Winston H.

    2014-01-01

    Unconstrained video recognition and Deep Convolution Network (DCN) are two active topics in computer vision recently. In this work, we apply DCNs as frame-based recognizers for video recognition. Our preliminary studies, however, show that video corpora with complete ground truth are usually not large and diverse enough to learn a robust model. The networks trained directly on the video data set suffer from significant overfitting and have poor recognition rate on the test set. The same lack-...

  14. Applicability of the Fourier convolution theorem to the analysis of late-type stellar spectra

    International Nuclear Information System (INIS)

    Bruning, D.H.

    1981-01-01

    Solar flux and intensity measurements were obtained at Sacramento Peak Observatory to test the validity of the Fourier convolution method as a means of analyzing the spectral line shapes of late-type stars. Analysis of six iron lines near 6200A shows that, in general, the convolution method is not a suitable approximation for the calculation of the flux profile. The convolution method does reasonably reproduce the line shape for some lines which appear not to vary across the disk of the sun, but does not properly calculate the central line depth of these lines. Even if a central depth correction could be found, it is difficult to predict, especially for stars other than the sun, which lines have nearly constant shapes and could be used with the convolution method. Therefore, explicit disk integrations are promoted as the only reliable method of spectral line analysis for late-type stars. Several methods of performing the disk integration are investigated. Although the Abt (1957) prescription appears suitable for the limited case studied, methods using annuli of equal area, equal flux, or equal width (Soberblom, 1980) are considered better models. The model that is the easiest to use and most efficient computationally is the equal area model. Model atmosphere calculations yield values for the microturbulence and macroturbulence similar to those derived by observers. Since the depth dependence of the microturbulence is ignored in the calculations, the intensity profiles at disk center and the limb do not match the observed intensity profiles with only one set of velocity parameters. Use of these incorrectly calculated intensity profiles in the integration procedure to obtain the flux profile leads to incorrect estimates of the solar macroturbulence

  15. Chinese Sentence Classification Based on Convolutional Neural Network

    Science.gov (United States)

    Gu, Chengwei; Wu, Ming; Zhang, Chuang

    2017-10-01

    Sentence classification is one of the significant issues in Natural Language Processing (NLP). Feature extraction is often regarded as the key point for natural language processing. Traditional ways based on machine learning can not take high level features into consideration, such as Naive Bayesian Model. The neural network for sentence classification can make use of contextual information to achieve greater results in sentence classification tasks. In this paper, we focus on classifying Chinese sentences. And the most important is that we post a novel architecture of Convolutional Neural Network (CNN) to apply on Chinese sentence classification. In particular, most of the previous methods often use softmax classifier for prediction, we embed a linear support vector machine to substitute softmax in the deep neural network model, minimizing a margin-based loss to get a better result. And we use tanh as an activation function, instead of ReLU. The CNN model improve the result of Chinese sentence classification tasks. Experimental results on the Chinese news title database validate the effectiveness of our model.

  16. High Order Tensor Formulation for Convolutional Sparse Coding

    KAUST Repository

    Bibi, Adel Aamer; Ghanem, Bernard

    2017-01-01

    Convolutional sparse coding (CSC) has gained attention for its successful role as a reconstruction and a classification tool in the computer vision and machine learning community. Current CSC methods can only reconstruct singlefeature 2D images

  17. Adversarial training and dilated convolutions for brain MRI segmentation

    NARCIS (Netherlands)

    Moeskops, P.; Veta, M.; Lafarge, M.W.; Eppenhof, K.A.J.; Pluim, J.P.W.

    2017-01-01

    Convolutional neural networks (CNNs) have been applied to various automatic image segmentation tasks in medical image analysis, including brain MRI segmentation. Generative adversarial networks have recently gained popularity because of their power in generating images that are difficult to

  18. Classification of urine sediment based on convolution neural network

    Science.gov (United States)

    Pan, Jingjing; Jiang, Cunbo; Zhu, Tiantian

    2018-04-01

    By designing a new convolution neural network framework, this paper breaks the constraints of the original convolution neural network framework requiring large training samples and samples of the same size. Move and cropping the input images, generate the same size of the sub-graph. And then, the generated sub-graph uses the method of dropout, increasing the diversity of samples and preventing the fitting generation. Randomly select some proper subset in the sub-graphic set and ensure that the number of elements in the proper subset is same and the proper subset is not the same. The proper subsets are used as input layers for the convolution neural network. Through the convolution layer, the pooling, the full connection layer and output layer, we can obtained the classification loss rate of test set and training set. In the red blood cells, white blood cells, calcium oxalate crystallization classification experiment, the classification accuracy rate of 97% or more.

  19. Convolution of second order linear recursive sequences II.

    Directory of Open Access Journals (Sweden)

    Szakács Tamás

    2017-12-01

    Full Text Available We continue the investigation of convolutions of second order linear recursive sequences (see the first part in [1]. In this paper, we focus on the case when the characteristic polynomials of the sequences have common root.

  20. FPGA-based digital convolution for wireless applications

    CERN Document Server

    Guan, Lei

    2017-01-01

    This book presents essential perspectives on digital convolutions in wireless communications systems and illustrates their corresponding efficient real-time field-programmable gate array (FPGA) implementations. Covering these digital convolutions from basic concept to vivid simulation/illustration, the book is also supplemented with MS PowerPoint presentations to aid in comprehension. FPGAs or generic all programmable devices will soon become widespread, serving as the “brains” of all types of real-time smart signal processing systems, like smart networks, smart homes and smart cities. The book examines digital convolution by bringing together the following main elements: the fundamental theory behind the mathematical formulae together with corresponding physical phenomena; virtualized algorithm simulation together with benchmark real-time FPGA implementations; and detailed, state-of-the-art case studies on wireless applications, including popular linear convolution in digital front ends (DFEs); nonlinear...

  1. Automated Detection of Obstructive Sleep Apnea Events from a Single-Lead Electrocardiogram Using a Convolutional Neural Network.

    Science.gov (United States)

    Urtnasan, Erdenebayar; Park, Jong-Uk; Joo, Eun-Yeon; Lee, Kyoung-Joung

    2018-04-23

    In this study, we propose a method for the automated detection of obstructive sleep apnea (OSA) from a single-lead electrocardiogram (ECG) using a convolutional neural network (CNN). A CNN model was designed with six optimized convolution layers including activation, pooling, and dropout layers. One-dimensional (1D) convolution, rectified linear units (ReLU), and max pooling were applied to the convolution, activation, and pooling layers, respectively. For training and evaluation of the CNN model, a single-lead ECG dataset was collected from 82 subjects with OSA and was divided into training (including data from 63 patients with 34,281 events) and testing (including data from 19 patients with 8571 events) datasets. Using this CNN model, a precision of 0.99%, a recall of 0.99%, and an F 1 -score of 0.99% were attained with the training dataset; these values were all 0.96% when the CNN was applied to the testing dataset. These results show that the proposed CNN model can be used to detect OSA accurately on the basis of a single-lead ECG. Ultimately, this CNN model may be used as a screening tool for those suspected to suffer from OSA.

  2. Traffic sign recognition with deep convolutional neural networks

    OpenAIRE

    Karamatić, Boris

    2016-01-01

    The problem of detection and recognition of traffic signs is becoming an important problem when it comes to the development of self driving cars and advanced driver assistance systems. In this thesis we will develop a system for detection and recognition of traffic signs. For the problem of detection we will use aggregate channel features and for the problem of recognition we will use a deep convolutional neural network. We will describe how convolutional neural networks work, how they are co...

  3. Convolutional Codes with Maximum Column Sum Rank for Network Streaming

    OpenAIRE

    Mahmood, Rafid; Badr, Ahmed; Khisti, Ashish

    2015-01-01

    The column Hamming distance of a convolutional code determines the error correction capability when streaming over a class of packet erasure channels. We introduce a metric known as the column sum rank, that parallels column Hamming distance when streaming over a network with link failures. We prove rank analogues of several known column Hamming distance properties and introduce a new family of convolutional codes that maximize the column sum rank up to the code memory. Our construction invol...

  4. Efficient and Invariant Convolutional Neural Networks for Dense Prediction

    OpenAIRE

    Gao, Hongyang; Ji, Shuiwang

    2017-01-01

    Convolutional neural networks have shown great success on feature extraction from raw input data such as images. Although convolutional neural networks are invariant to translations on the inputs, they are not invariant to other transformations, including rotation and flip. Recent attempts have been made to incorporate more invariance in image recognition applications, but they are not applicable to dense prediction tasks, such as image segmentation. In this paper, we propose a set of methods...

  5. Research of convolutional neural networks for traffic sign recognition

    OpenAIRE

    Stadalnikas, Kasparas

    2017-01-01

    In this thesis the convolutional neural networks application for traffic sign recognition is analyzed. Thesis describes the basic operations, techniques that are commonly used to apply in the image classification using convolutional neural networks. Also, this paper describes the data sets used for traffic sign recognition, their problems affecting the final training results. The paper reviews most popular existing technologies – frameworks for developing the solution for traffic sign recogni...

  6. On the Fresnel sine integral and the convolution

    Directory of Open Access Journals (Sweden)

    Adem Kılıçman

    2003-01-01

    Full Text Available The Fresnel sine integral S(x, the Fresnel cosine integral C(x, and the associated functions S+(x, S−(x, C+(x, and C−(x are defined as locally summable functions on the real line. Some convolutions and neutrix convolutions of the Fresnel sine integral and its associated functions with x+r, xr are evaluated.

  7. Convolution-based estimation of organ dose in tube current modulated CT

    Science.gov (United States)

    Tian, Xiaoyu; Segars, W. Paul; Dixon, Robert L.; Samei, Ehsan

    2016-05-01

    Estimating organ dose for clinical patients requires accurate modeling of the patient anatomy and the dose field of the CT exam. The modeling of patient anatomy can be achieved using a library of representative computational phantoms (Samei et al 2014 Pediatr. Radiol. 44 460-7). The modeling of the dose field can be challenging for CT exams performed with a tube current modulation (TCM) technique. The purpose of this work was to effectively model the dose field for TCM exams using a convolution-based method. A framework was further proposed for prospective and retrospective organ dose estimation in clinical practice. The study included 60 adult patients (age range: 18-70 years, weight range: 60-180 kg). Patient-specific computational phantoms were generated based on patient CT image datasets. A previously validated Monte Carlo simulation program was used to model a clinical CT scanner (SOMATOM Definition Flash, Siemens Healthcare, Forchheim, Germany). A practical strategy was developed to achieve real-time organ dose estimation for a given clinical patient. CTDIvol-normalized organ dose coefficients ({{h}\\text{Organ}} ) under constant tube current were estimated and modeled as a function of patient size. Each clinical patient in the library was optimally matched to another computational phantom to obtain a representation of organ location/distribution. The patient organ distribution was convolved with a dose distribution profile to generate {{≤ft(\\text{CTD}{{\\text{I}}\\text{vol}}\\right)}\\text{organ, \\text{convolution}}} values that quantified the regional dose field for each organ. The organ dose was estimated by multiplying {{≤ft(\\text{CTD}{{\\text{I}}\\text{vol}}\\right)}\\text{organ, \\text{convolution}}} with the organ dose coefficients ({{h}\\text{Organ}} ). To validate the accuracy of this dose estimation technique, the organ dose of the original clinical patient was estimated using Monte Carlo program with TCM profiles explicitly modeled. The

  8. Deep Convolutional Neural Networks for Classifying Body Constitution Based on Face Image.

    Science.gov (United States)

    Huan, Er-Yang; Wen, Gui-Hua; Zhang, Shi-Jun; Li, Dan-Yang; Hu, Yang; Chang, Tian-Yuan; Wang, Qing; Huang, Bing-Lin

    2017-01-01

    Body constitution classification is the basis and core content of traditional Chinese medicine constitution research. It is to extract the relevant laws from the complex constitution phenomenon and finally build the constitution classification system. Traditional identification methods have the disadvantages of inefficiency and low accuracy, for instance, questionnaires. This paper proposed a body constitution recognition algorithm based on deep convolutional neural network, which can classify individual constitution types according to face images. The proposed model first uses the convolutional neural network to extract the features of face image and then combines the extracted features with the color features. Finally, the fusion features are input to the Softmax classifier to get the classification result. Different comparison experiments show that the algorithm proposed in this paper can achieve the accuracy of 65.29% about the constitution classification. And its performance was accepted by Chinese medicine practitioners.

  9. A Novel Image Tag Completion Method Based on Convolutional Neural Transformation

    KAUST Repository

    Geng, Yanyan; Zhang, Guohui; Li, Weizhi; Gu, Yi; Liang, Ru-Ze; Liang, Gaoyuan; Wang, Jingbin; Wu, Yanbin; Patil, Nitin; Wang, Jing-Yan

    2017-01-01

    In the problems of image retrieval and annotation, complete textual tag lists of images play critical roles. However, in real-world applications, the image tags are usually incomplete, thus it is important to learn the complete tags for images. In this paper, we study the problem of image tag complete and proposed a novel method for this problem based on a popular image representation method, convolutional neural network (CNN). The method estimates the complete tags from the convolutional filtering outputs of images based on a linear predictor. The CNN parameters, linear predictor, and the complete tags are learned jointly by our method. We build a minimization problem to encourage the consistency between the complete tags and the available incomplete tags, reduce the estimation error, and reduce the model complexity. An iterative algorithm is developed to solve the minimization problem. Experiments over benchmark image data sets show its effectiveness.

  10. A Novel Image Tag Completion Method Based on Convolutional Neural Transformation

    KAUST Repository

    Geng, Yanyan

    2017-10-24

    In the problems of image retrieval and annotation, complete textual tag lists of images play critical roles. However, in real-world applications, the image tags are usually incomplete, thus it is important to learn the complete tags for images. In this paper, we study the problem of image tag complete and proposed a novel method for this problem based on a popular image representation method, convolutional neural network (CNN). The method estimates the complete tags from the convolutional filtering outputs of images based on a linear predictor. The CNN parameters, linear predictor, and the complete tags are learned jointly by our method. We build a minimization problem to encourage the consistency between the complete tags and the available incomplete tags, reduce the estimation error, and reduce the model complexity. An iterative algorithm is developed to solve the minimization problem. Experiments over benchmark image data sets show its effectiveness.

  11. REAL-TIME VIDEO SCALING BASED ON CONVOLUTION NEURAL NETWORK ARCHITECTURE

    Directory of Open Access Journals (Sweden)

    S Safinaz

    2017-08-01

    Full Text Available In recent years, video super resolution techniques becomes mandatory requirements to get high resolution videos. Many super resolution techniques researched but still video super resolution or scaling is a vital challenge. In this paper, we have presented a real-time video scaling based on convolution neural network architecture to eliminate the blurriness in the images and video frames and to provide better reconstruction quality while scaling of large datasets from lower resolution frames to high resolution frames. We compare our outcomes with multiple exiting algorithms. Our extensive results of proposed technique RemCNN (Reconstruction error minimization Convolution Neural Network shows that our model outperforms the existing technologies such as bicubic, bilinear, MCResNet and provide better reconstructed motioning images and video frames. The experimental results shows that our average PSNR result is 47.80474 considering upscale-2, 41.70209 for upscale-3 and 36.24503 for upscale-4 for Myanmar dataset which is very high in contrast to other existing techniques. This results proves our proposed model real-time video scaling based on convolution neural network architecture’s high efficiency and better performance.

  12. Chromatin accessibility prediction via convolutional long short-term memory networks with k-mer embedding.

    Science.gov (United States)

    Min, Xu; Zeng, Wanwen; Chen, Ning; Chen, Ting; Jiang, Rui

    2017-07-15

    Experimental techniques for measuring chromatin accessibility are expensive and time consuming, appealing for the development of computational approaches to predict open chromatin regions from DNA sequences. Along this direction, existing methods fall into two classes: one based on handcrafted k -mer features and the other based on convolutional neural networks. Although both categories have shown good performance in specific applications thus far, there still lacks a comprehensive framework to integrate useful k -mer co-occurrence information with recent advances in deep learning. We fill this gap by addressing the problem of chromatin accessibility prediction with a convolutional Long Short-Term Memory (LSTM) network with k -mer embedding. We first split DNA sequences into k -mers and pre-train k -mer embedding vectors based on the co-occurrence matrix of k -mers by using an unsupervised representation learning approach. We then construct a supervised deep learning architecture comprised of an embedding layer, three convolutional layers and a Bidirectional LSTM (BLSTM) layer for feature learning and classification. We demonstrate that our method gains high-quality fixed-length features from variable-length sequences and consistently outperforms baseline methods. We show that k -mer embedding can effectively enhance model performance by exploring different embedding strategies. We also prove the efficacy of both the convolution and the BLSTM layers by comparing two variations of the network architecture. We confirm the robustness of our model to hyper-parameters by performing sensitivity analysis. We hope our method can eventually reinforce our understanding of employing deep learning in genomic studies and shed light on research regarding mechanisms of chromatin accessibility. The source code can be downloaded from https://github.com/minxueric/ismb2017_lstm . tingchen@tsinghua.edu.cn or ruijiang@tsinghua.edu.cn. Supplementary materials are available at

  13. Metaheuristic Algorithms for Convolution Neural Network.

    Science.gov (United States)

    Rere, L M Rasdi; Fanany, Mohamad Ivan; Arymurthy, Aniati Murni

    2016-01-01

    A typical modern optimization technique is usually either heuristic or metaheuristic. This technique has managed to solve some optimization problems in the research area of science, engineering, and industry. However, implementation strategy of metaheuristic for accuracy improvement on convolution neural networks (CNN), a famous deep learning method, is still rarely investigated. Deep learning relates to a type of machine learning technique, where its aim is to move closer to the goal of artificial intelligence of creating a machine that could successfully perform any intellectual tasks that can be carried out by a human. In this paper, we propose the implementation strategy of three popular metaheuristic approaches, that is, simulated annealing, differential evolution, and harmony search, to optimize CNN. The performances of these metaheuristic methods in optimizing CNN on classifying MNIST and CIFAR dataset were evaluated and compared. Furthermore, the proposed methods are also compared with the original CNN. Although the proposed methods show an increase in the computation time, their accuracy has also been improved (up to 7.14 percent).

  14. Metaheuristic Algorithms for Convolution Neural Network

    Directory of Open Access Journals (Sweden)

    L. M. Rasdi Rere

    2016-01-01

    Full Text Available A typical modern optimization technique is usually either heuristic or metaheuristic. This technique has managed to solve some optimization problems in the research area of science, engineering, and industry. However, implementation strategy of metaheuristic for accuracy improvement on convolution neural networks (CNN, a famous deep learning method, is still rarely investigated. Deep learning relates to a type of machine learning technique, where its aim is to move closer to the goal of artificial intelligence of creating a machine that could successfully perform any intellectual tasks that can be carried out by a human. In this paper, we propose the implementation strategy of three popular metaheuristic approaches, that is, simulated annealing, differential evolution, and harmony search, to optimize CNN. The performances of these metaheuristic methods in optimizing CNN on classifying MNIST and CIFAR dataset were evaluated and compared. Furthermore, the proposed methods are also compared with the original CNN. Although the proposed methods show an increase in the computation time, their accuracy has also been improved (up to 7.14 percent.

  15. Microaneurysm detection using fully convolutional neural networks.

    Science.gov (United States)

    Chudzik, Piotr; Majumdar, Somshubra; Calivá, Francesco; Al-Diri, Bashir; Hunter, Andrew

    2018-05-01

    Diabetic retinopathy is a microvascular complication of diabetes that can lead to sight loss if treated not early enough. Microaneurysms are the earliest clinical signs of diabetic retinopathy. This paper presents an automatic method for detecting microaneurysms in fundus photographies. A novel patch-based fully convolutional neural network with batch normalization layers and Dice loss function is proposed. Compared to other methods that require up to five processing stages, it requires only three. Furthermore, to the best of the authors' knowledge, this is the first paper that shows how to successfully transfer knowledge between datasets in the microaneurysm detection domain. The proposed method was evaluated using three publicly available and widely used datasets: E-Ophtha, DIARETDB1, and ROC. It achieved better results than state-of-the-art methods using the FROC metric. The proposed algorithm accomplished highest sensitivities for low false positive rates, which is particularly important for screening purposes. Performance, simplicity, and robustness of the proposed method demonstrates its suitability for diabetic retinopathy screening applications. Copyright © 2018 Elsevier B.V. All rights reserved.

  16. Real-time hybrid simulation using the convolution integral method

    International Nuclear Information System (INIS)

    Kim, Sung Jig; Christenson, Richard E; Wojtkiewicz, Steven F; Johnson, Erik A

    2011-01-01

    This paper proposes a real-time hybrid simulation method that will allow complex systems to be tested within the hybrid test framework by employing the convolution integral (CI) method. The proposed CI method is potentially transformative for real-time hybrid simulation. The CI method can allow real-time hybrid simulation to be conducted regardless of the size and complexity of the numerical model and for numerical stability to be ensured in the presence of high frequency responses in the simulation. This paper presents the general theory behind the proposed CI method and provides experimental verification of the proposed method by comparing the CI method to the current integration time-stepping (ITS) method. Real-time hybrid simulation is conducted in the Advanced Hazard Mitigation Laboratory at the University of Connecticut. A seismically excited two-story shear frame building with a magneto-rheological (MR) fluid damper is selected as the test structure to experimentally validate the proposed method. The building structure is numerically modeled and simulated, while the MR damper is physically tested. Real-time hybrid simulation using the proposed CI method is shown to provide accurate results

  17. Bilinear Convolutional Neural Networks for Fine-grained Visual Recognition.

    Science.gov (United States)

    Lin, Tsung-Yu; RoyChowdhury, Aruni; Maji, Subhransu

    2017-07-04

    We present a simple and effective architecture for fine-grained recognition called Bilinear Convolutional Neural Networks (B-CNNs). These networks represent an image as a pooled outer product of features derived from two CNNs and capture localized feature interactions in a translationally invariant manner. B-CNNs are related to orderless texture representations built on deep features but can be trained in an end-to-end manner. Our most accurate model obtains 84.1%, 79.4%, 84.5% and 91.3% per-image accuracy on the Caltech-UCSD birds [66], NABirds [63], FGVC aircraft [42], and Stanford cars [33] dataset respectively and runs at 30 frames-per-second on a NVIDIA Titan X GPU. We then present a systematic analysis of these networks and show that (1) the bilinear features are highly redundant and can be reduced by an order of magnitude in size without significant loss in accuracy, (2) are also effective for other image classification tasks such as texture and scene recognition, and (3) can be trained from scratch on the ImageNet dataset offering consistent improvements over the baseline architecture. Finally, we present visualizations of these models on various datasets using top activations of neural units and gradient-based inversion techniques. The source code for the complete system is available at http://vis-www.cs.umass.edu/bcnn.

  18. Image reconstruction in computerized tomography using the convolution method

    International Nuclear Information System (INIS)

    Oliveira Rebelo, A.M. de.

    1984-03-01

    In the present work an algoritin was derived, using the analytical convolution method (filtered back-projection) for two-dimensional or three-dimensional image reconstruction in computerized tomography applied to non-destructive testing and to the medical use. This mathematical model is based on the analytical Fourier transform method for image reconstruction. This model consists of a discontinuous system formed by an NxN array of cells (pixels). The attenuation in the object under study of a colimated gamma ray beam has been determined for various positions and incidence angles (projections) in terms of the interaction of the beam with the intercepted pixels. The contribution of each pixel to beam attenuation was determined using the weight function W ij which was used for simulated tests. Simulated tests using standard objects with attenuation coefficients in the range of 0,2 to 0,7 cm -1 were carried out using cell arrays of up to 25x25. One application was carried out in the medical area simulating image reconstruction of an arm phantom with attenuation coefficients in the range of 0,2 to 0,5 cm -1 using cell arrays of 41x41. The simulated results show that, in objects with a great number of interfaces and great variations of attenuation coefficients at these interfaces, a good reconstruction is obtained with the number of projections equal to the reconstruction matrix dimension. A good reconstruction is otherwise obtained with fewer projections. (author) [pt

  19. Reconstruction of Micropattern Detector Signals using Convolutional Neural Networks

    Science.gov (United States)

    Flekova, L.; Schott, M.

    2017-10-01

    Micropattern gaseous detector (MPGD) technologies, such as GEMs or MicroMegas, are particularly suitable for precision tracking and triggering in high rate environments. Given their relatively low production costs, MPGDs are an exemplary candidate for the next generation of particle detectors. Having acknowledged these advantages, both the ATLAS and CMS collaborations at the LHC are exploiting these new technologies for their detector upgrade programs in the coming years. When MPGDs are utilized for triggering purposes, the measured signals need to be precisely reconstructed within less than 200 ns, which can be achieved by the usage of FPGAs. In this work, we present a novel approach to identify reconstructed signals, their timing and the corresponding spatial position on the detector. In particular, we study the effect of noise and dead readout strips on the reconstruction performance. Our approach leverages the potential of convolutional neural network (CNNs), which have recently manifested an outstanding performance in a range of modeling tasks. The proposed neural network architecture of our CNN is designed simply enough, so that it can be modeled directly by an FPGA and thus provide precise information on reconstructed signals already in trigger level.

  20. Convolutional neural network with transfer learning for rice type classification

    Science.gov (United States)

    Patel, Vaibhav Amit; Joshi, Manjunath V.

    2018-04-01

    Presently, rice type is identified manually by humans, which is time consuming and error prone. Therefore, there is a need to do this by machine which makes it faster with greater accuracy. This paper proposes a deep learning based method for classification of rice types. We propose two methods to classify the rice types. In the first method, we train a deep convolutional neural network (CNN) using the given segmented rice images. In the second method, we train a combination of a pretrained VGG16 network and the proposed method, while using transfer learning in which the weights of a pretrained network are used to achieve better accuracy. Our approach can also be used for classification of rice grain as broken or fine. We train a 5-class model for classifying rice types using 4000 training images and another 2- class model for the classification of broken and normal rice using 1600 training images. We observe that despite having distinct rice images, our architecture, pretrained on ImageNet data boosts classification accuracy significantly.

  1. Transfer Learning with Convolutional Neural Networks for SAR Ship Recognition

    Science.gov (United States)

    Zhang, Di; Liu, Jia; Heng, Wang; Ren, Kaijun; Song, Junqiang

    2018-03-01

    Ship recognition is the backbone of marine surveillance systems. Recent deep learning methods, e.g. Convolutional Neural Networks (CNNs), have shown high performance for optical images. Learning CNNs, however, requires a number of annotated samples to estimate numerous model parameters, which prevents its application to Synthetic Aperture Radar (SAR) images due to the limited annotated training samples. Transfer learning has been a promising technique for applications with limited data. To this end, a novel SAR ship recognition method based on CNNs with transfer learning has been developed. In this work, we firstly start with a CNNs model that has been trained in advance on Moving and Stationary Target Acquisition and Recognition (MSTAR) database. Next, based on the knowledge gained from this image recognition task, we fine-tune the CNNs on a new task to recognize three types of ships in the OpenSARShip database. The experimental results show that our proposed approach can obviously increase the recognition rate comparing with the result of merely applying CNNs. In addition, compared to existing methods, the proposed method proves to be very competitive and can learn discriminative features directly from training data instead of requiring pre-specification or pre-selection manually.

  2. Bidirectional-Convolutional LSTM Based Spectral-Spatial Feature Learning for Hyperspectral Image Classification

    Directory of Open Access Journals (Sweden)

    Qingshan Liu

    2017-12-01

    Full Text Available This paper proposes a novel deep learning framework named bidirectional-convolutional long short term memory (Bi-CLSTM network to automatically learn the spectral-spatial features from hyperspectral images (HSIs. In the network, the issue of spectral feature extraction is considered as a sequence learning problem, and a recurrent connection operator across the spectral domain is used to address it. Meanwhile, inspired from the widely used convolutional neural network (CNN, a convolution operator across the spatial domain is incorporated into the network to extract the spatial feature. In addition, to sufficiently capture the spectral information, a bidirectional recurrent connection is proposed. In the classification phase, the learned features are concatenated into a vector and fed to a Softmax classifier via a fully-connected operator. To validate the effectiveness of the proposed Bi-CLSTM framework, we compare it with six state-of-the-art methods, including the popular 3D-CNN model, on three widely used HSIs (i.e., Indian Pines, Pavia University, and Kennedy Space Center. The obtained results show that Bi-CLSTM can improve the classification performance by almost 1.5 % as compared to 3D-CNN.

  3. Forecasting short-term data center network traffic load with convolutional neural networks

    Science.gov (United States)

    Ordozgoiti, Bruno; Gómez-Canaval, Sandra

    2018-01-01

    Efficient resource management in data centers is of central importance to content service providers as 90 percent of the network traffic is expected to go through them in the coming years. In this context we propose the use of convolutional neural networks (CNNs) to forecast short-term changes in the amount of traffic crossing a data center network. This value is an indicator of virtual machine activity and can be utilized to shape the data center infrastructure accordingly. The behaviour of network traffic at the seconds scale is highly chaotic and therefore traditional time-series-analysis approaches such as ARIMA fail to obtain accurate forecasts. We show that our convolutional neural network approach can exploit the non-linear regularities of network traffic, providing significant improvements with respect to the mean absolute and standard deviation of the data, and outperforming ARIMA by an increasingly significant margin as the forecasting granularity is above the 16-second resolution. In order to increase the accuracy of the forecasting model, we exploit the architecture of the CNNs using multiresolution input distributed among separate channels of the first convolutional layer. We validate our approach with an extensive set of experiments using a data set collected at the core network of an Internet Service Provider over a period of 5 months, totalling 70 days of traffic at the one-second resolution. PMID:29408936

  4. Forecasting short-term data center network traffic load with convolutional neural networks.

    Science.gov (United States)

    Mozo, Alberto; Ordozgoiti, Bruno; Gómez-Canaval, Sandra

    2018-01-01

    Efficient resource management in data centers is of central importance to content service providers as 90 percent of the network traffic is expected to go through them in the coming years. In this context we propose the use of convolutional neural networks (CNNs) to forecast short-term changes in the amount of traffic crossing a data center network. This value is an indicator of virtual machine activity and can be utilized to shape the data center infrastructure accordingly. The behaviour of network traffic at the seconds scale is highly chaotic and therefore traditional time-series-analysis approaches such as ARIMA fail to obtain accurate forecasts. We show that our convolutional neural network approach can exploit the non-linear regularities of network traffic, providing significant improvements with respect to the mean absolute and standard deviation of the data, and outperforming ARIMA by an increasingly significant margin as the forecasting granularity is above the 16-second resolution. In order to increase the accuracy of the forecasting model, we exploit the architecture of the CNNs using multiresolution input distributed among separate channels of the first convolutional layer. We validate our approach with an extensive set of experiments using a data set collected at the core network of an Internet Service Provider over a period of 5 months, totalling 70 days of traffic at the one-second resolution.

  5. Joint multiple fully connected convolutional neural network with extreme learning machine for hepatocellular carcinoma nuclei grading.

    Science.gov (United States)

    Li, Siqi; Jiang, Huiyan; Pang, Wenbo

    2017-05-01

    Accurate cell grading of cancerous tissue pathological image is of great importance in medical diagnosis and treatment. This paper proposes a joint multiple fully connected convolutional neural network with extreme learning machine (MFC-CNN-ELM) architecture for hepatocellular carcinoma (HCC) nuclei grading. First, in preprocessing stage, each grayscale image patch with the fixed size is obtained using center-proliferation segmentation (CPS) method and the corresponding labels are marked under the guidance of three pathologists. Next, a multiple fully connected convolutional neural network (MFC-CNN) is designed to extract the multi-form feature vectors of each input image automatically, which considers multi-scale contextual information of deep layer maps sufficiently. After that, a convolutional neural network extreme learning machine (CNN-ELM) model is proposed to grade HCC nuclei. Finally, a back propagation (BP) algorithm, which contains a new up-sample method, is utilized to train MFC-CNN-ELM architecture. The experiment comparison results demonstrate that our proposed MFC-CNN-ELM has superior performance compared with related works for HCC nuclei grading. Meanwhile, external validation using ICPR 2014 HEp-2 cell dataset shows the good generalization of our MFC-CNN-ELM architecture. Copyright © 2017 Elsevier Ltd. All rights reserved.

  6. A pre-trained convolutional neural network based method for thyroid nodule diagnosis.

    Science.gov (United States)

    Ma, Jinlian; Wu, Fa; Zhu, Jiang; Xu, Dong; Kong, Dexing

    2017-01-01

    In ultrasound images, most thyroid nodules are in heterogeneous appearances with various internal components and also have vague boundaries, so it is difficult for physicians to discriminate malignant thyroid nodules from benign ones. In this study, we propose a hybrid method for thyroid nodule diagnosis, which is a fusion of two pre-trained convolutional neural networks (CNNs) with different convolutional layers and fully-connected layers. Firstly, the two networks pre-trained with ImageNet database are separately trained. Secondly, we fuse feature maps learned by trained convolutional filters, pooling and normalization operations of the two CNNs. Finally, with the fused feature maps, a softmax classifier is used to diagnose thyroid nodules. The proposed method is validated on 15,000 ultrasound images collected from two local hospitals. Experiment results show that the proposed CNN based methods can accurately and effectively diagnose thyroid nodules. In addition, the fusion of the two CNN based models lead to significant performance improvement, with an accuracy of 83.02%±0.72%. These demonstrate the potential clinical applications of this method. Copyright © 2016 Elsevier B.V. All rights reserved.

  7. Design and Implementation of Convolutional Encoder and Viterbi Decoder Using FPGA.

    Directory of Open Access Journals (Sweden)

    Riham Ali Zbaid

    2018-01-01

    Full Text Available Keeping  the  fineness of data is the most significant thing in communication.There are many factors that affect the accuracy of the data when it is transmitted over the communication channel such as noise etc. to overcome these effects are encoding channels encryption.In this paper is used for one type of channel coding is convolutional codes. Convolution encoding is a Forward Error Correction (FEC method used in incessant one-way and real time communication links .It can offer a great development in the error bit rates so that small, low energy, and devices cheap transmission when used in applications such as satellites. In this paper highlight the design, simulation and implementation of convolution encoder and Viterbi decoder by using MATLAB- program (2011. SIMULINK HDL coder is used to convert MATLAB-SIMULINK models to VHDL using plates Altera Cyclone II code DE2-70. Simulation and evaluation of the implementation of the results coincided with the results of the design show the coinciding with the designed results.

  8. Joint Multi-scale Convolution Neural Network for Scene Classification of High Resolution Remote Sensing Imagery

    Directory of Open Access Journals (Sweden)

    ZHENG Zhuo

    2018-05-01

    Full Text Available High resolution remote sensing imagery scene classification is important for automatic complex scene recognition, which is the key technology for military and disaster relief, etc. In this paper, we propose a novel joint multi-scale convolution neural network (JMCNN method using a limited amount of image data for high resolution remote sensing imagery scene classification. Different from traditional convolutional neural network, the proposed JMCNN is an end-to-end training model with joint enhanced high-level feature representation, which includes multi-channel feature extractor, joint multi-scale feature fusion and Softmax classifier. Multi-channel and scale convolutional extractors are used to extract scene middle features, firstly. Then, in order to achieve enhanced high-level feature representation in a limit dataset, joint multi-scale feature fusion is proposed to combine multi-channel and scale features using two feature fusions. Finally, enhanced high-level feature representation can be used for classification by Softmax. Experiments were conducted using two limit public UCM and SIRI datasets. Compared to state-of-the-art methods, the JMCNN achieved improved performance and great robustness with average accuracies of 89.3% and 88.3% on the two datasets.

  9. A quantum algorithm for Viterbi decoding of classical convolutional codes

    Science.gov (United States)

    Grice, Jon R.; Meyer, David A.

    2015-07-01

    We present a quantum Viterbi algorithm (QVA) with better than classical performance under certain conditions. In this paper, the proposed algorithm is applied to decoding classical convolutional codes, for instance, large constraint length and short decode frames . Other applications of the classical Viterbi algorithm where is large (e.g., speech processing) could experience significant speedup with the QVA. The QVA exploits the fact that the decoding trellis is similar to the butterfly diagram of the fast Fourier transform, with its corresponding fast quantum algorithm. The tensor-product structure of the butterfly diagram corresponds to a quantum superposition that we show can be efficiently prepared. The quantum speedup is possible because the performance of the QVA depends on the fanout (number of possible transitions from any given state in the hidden Markov model) which is in general much less than . The QVA constructs a superposition of states which correspond to all legal paths through the decoding lattice, with phase as a function of the probability of the path being taken given received data. A specialized amplitude amplification procedure is applied one or more times to recover a superposition where the most probable path has a high probability of being measured.

  10. On the Relationship between Visual Attributes and Convolutional Networks

    KAUST Repository

    Castillo, Victor; Ghanem, Bernard; Niebles, Juan Carlos

    2015-01-01

    One of the cornerstone principles of deep models is their abstraction capacity, i.e. their ability to learn abstract concepts from ‘simpler’ ones. Through extensive experiments, we characterize the nature of the relationship between abstract concepts (specifically objects in images) learned by popular and high performing convolutional networks (conv-nets) and established mid-level representations used in computer vision (specifically semantic visual attributes). We focus on attributes due to their impact on several applications, such as object description, retrieval and mining, and active (and zero-shot) learning. Among the findings we uncover, we show empirical evidence of the existence of Attribute Centric Nodes (ACNs) within a conv-net, which is trained to recognize objects (not attributes) in images. These special conv-net nodes (1) collectively encode information pertinent to visual attribute representation and discrimination, (2) are unevenly and sparsely distribution across all layers of the conv-net, and (3) play an important role in conv-net based object recognition.

  11. On the Relationship between Visual Attributes and Convolutional Networks

    KAUST Repository

    Castillo, Victor

    2015-06-02

    One of the cornerstone principles of deep models is their abstraction capacity, i.e. their ability to learn abstract concepts from ‘simpler’ ones. Through extensive experiments, we characterize the nature of the relationship between abstract concepts (specifically objects in images) learned by popular and high performing convolutional networks (conv-nets) and established mid-level representations used in computer vision (specifically semantic visual attributes). We focus on attributes due to their impact on several applications, such as object description, retrieval and mining, and active (and zero-shot) learning. Among the findings we uncover, we show empirical evidence of the existence of Attribute Centric Nodes (ACNs) within a conv-net, which is trained to recognize objects (not attributes) in images. These special conv-net nodes (1) collectively encode information pertinent to visual attribute representation and discrimination, (2) are unevenly and sparsely distribution across all layers of the conv-net, and (3) play an important role in conv-net based object recognition.

  12. An improved multi-domain convolution tracking algorithm

    Science.gov (United States)

    Sun, Xin; Wang, Haiying; Zeng, Yingsen

    2018-04-01

    Along with the wide application of the Deep Learning in the field of Computer vision, Deep learning has become a mainstream direction in the field of object tracking. The tracking algorithm in this paper is based on the improved multidomain convolution neural network, and the VOT video set is pre-trained on the network by multi-domain training strategy. In the process of online tracking, the network evaluates candidate targets sampled from vicinity of the prediction target in the previous with Gaussian distribution, and the candidate target with the highest score is recognized as the prediction target of this frame. The Bounding Box Regression model is introduced to make the prediction target closer to the ground-truths target box of the test set. Grouping-update strategy is involved to extract and select useful update samples in each frame, which can effectively prevent over fitting. And adapt to changes in both target and environment. To improve the speed of the algorithm while maintaining the performance, the number of candidate target succeed in adjusting dynamically with the help of Self-adaption parameter Strategy. Finally, the algorithm is tested by OTB set, compared with other high-performance tracking algorithms, and the plot of success rate and the accuracy are drawn. which illustrates outstanding performance of the tracking algorithm in this paper.

  13. A Hierarchical Convolutional Neural Network for vesicle fusion event classification.

    Science.gov (United States)

    Li, Haohan; Mao, Yunxiang; Yin, Zhaozheng; Xu, Yingke

    2017-09-01

    Quantitative analysis of vesicle exocytosis and classification of different modes of vesicle fusion from the fluorescence microscopy are of primary importance for biomedical researches. In this paper, we propose a novel Hierarchical Convolutional Neural Network (HCNN) method to automatically identify vesicle fusion events in time-lapse Total Internal Reflection Fluorescence Microscopy (TIRFM) image sequences. Firstly, a detection and tracking method is developed to extract image patch sequences containing potential fusion events. Then, a Gaussian Mixture Model (GMM) is applied on each image patch of the patch sequence with outliers rejected for robust Gaussian fitting. By utilizing the high-level time-series intensity change features introduced by GMM and the visual appearance features embedded in some key moments of the fusion process, the proposed HCNN architecture is able to classify each candidate patch sequence into three classes: full fusion event, partial fusion event and non-fusion event. Finally, we validate the performance of our method on 9 challenging datasets that have been annotated by cell biologists, and our method achieves better performances when comparing with three previous methods. Copyright © 2017 Elsevier Ltd. All rights reserved.

  14. Classification and Segmentation of Satellite Orthoimagery Using Convolutional Neural Networks

    Directory of Open Access Journals (Sweden)

    Martin Längkvist

    2016-04-01

    Full Text Available The availability of high-resolution remote sensing (HRRS data has opened up the possibility for new interesting applications, such as per-pixel classification of individual objects in greater detail. This paper shows how a convolutional neural network (CNN can be applied to multispectral orthoimagery and a digital surface model (DSM of a small city for a full, fast and accurate per-pixel classification. The predicted low-level pixel classes are then used to improve the high-level segmentation. Various design choices of the CNN architecture are evaluated and analyzed. The investigated land area is fully manually labeled into five categories (vegetation, ground, roads, buildings and water, and the classification accuracy is compared to other per-pixel classification works on other land areas that have a similar choice of categories. The results of the full classification and segmentation on selected segments of the map show that CNNs are a viable tool for solving both the segmentation and object recognition task for remote sensing data.

  15. Pneumothorax detection in chest radiographs using convolutional neural networks

    Science.gov (United States)

    Blumenfeld, Aviel; Konen, Eli; Greenspan, Hayit

    2018-02-01

    This study presents a computer assisted diagnosis system for the detection of pneumothorax (PTX) in chest radiographs based on a convolutional neural network (CNN) for pixel classification. Using a pixel classification approach allows utilization of the texture information in the local environment of each pixel while training a CNN model on millions of training patches extracted from a relatively small dataset. The proposed system uses a pre-processing step of lung field segmentation to overcome the large variability in the input images coming from a variety of imaging sources and protocols. Using a CNN classification, suspected pixel candidates are extracted within each lung segment. A postprocessing step follows to remove non-physiological suspected regions and noisy connected components. The overall percentage of suspected PTX area was used as a robust global decision for the presence of PTX in each lung. The system was trained on a set of 117 chest x-ray images with ground truth segmentations of the PTX regions. The system was tested on a set of 86 images and reached diagnosis accuracy of AUC=0.95. Overall preliminary results are promising and indicate the growing ability of CAD based systems to detect findings in medical imaging on a clinical level accuracy.

  16. Vision-Based Fall Detection with Convolutional Neural Networks

    Directory of Open Access Journals (Sweden)

    Adrián Núñez-Marcos

    2017-01-01

    Full Text Available One of the biggest challenges in modern societies is the improvement of healthy aging and the support to older persons in their daily activities. In particular, given its social and economic impact, the automatic detection of falls has attracted considerable attention in the computer vision and pattern recognition communities. Although the approaches based on wearable sensors have provided high detection rates, some of the potential users are reluctant to wear them and thus their use is not yet normalized. As a consequence, alternative approaches such as vision-based methods have emerged. We firmly believe that the irruption of the Smart Environments and the Internet of Things paradigms, together with the increasing number of cameras in our daily environment, forms an optimal context for vision-based systems. Consequently, here we propose a vision-based solution using Convolutional Neural Networks to decide if a sequence of frames contains a person falling. To model the video motion and make the system scenario independent, we use optical flow images as input to the networks followed by a novel three-step training phase. Furthermore, our method is evaluated in three public datasets achieving the state-of-the-art results in all three of them.

  17. Innervation of the renal proximal convoluted tubule of the rat

    International Nuclear Information System (INIS)

    Barajas, L.; Powers, K.

    1989-01-01

    Experimental data suggest the proximal tubule as a major site of neurogenic influence on tubular function. The functional and anatomical axial heterogeneity of the proximal tubule prompted this study of the distribution of innervation sites along the early, mid, and late proximal convoluted tubule (PCT) of the rat. Serial section autoradiograms, with tritiated norepinephrine serving as a marker for monoaminergic nerves, were used in this study. Freehand clay models and graphic reconstructions of proximal tubules permitted a rough estimation of the location of the innervation sites along the PCT. In the subcapsular nephrons, the early PCT (first third) was devoid of innervation sites with most of the innervation occurring in the mid (middle third) and in the late (last third) PCT. Innervation sites were found in the early PCT in nephrons located deeper in the cortex. In juxtamedullary nephrons, innervation sites could be observed on the PCT as it left the glomerulus. This gradient of PCT innervation can be explained by the different tubulovascular relationships of nephrons at different levels of the cortex. The absence of innervation sites in the early PCT of subcapsular nephrons suggests that any influence of the renal nerves on the early PCT might be due to an effect of neurotransmitter released from renal nerves reaching the early PCT via the interstitium and/or capillaries

  18. Classification of breast cancer cytological specimen using convolutional neural network

    Science.gov (United States)

    Żejmo, Michał; Kowal, Marek; Korbicz, Józef; Monczak, Roman

    2017-01-01

    The paper presents a deep learning approach for automatic classification of breast tumors based on fine needle cytology. The main aim of the system is to distinguish benign from malignant cases based on microscopic images. Experiment was carried out on cytological samples derived from 50 patients (25 benign cases + 25 malignant cases) diagnosed in Regional Hospital in Zielona Góra. To classify microscopic images, we used convolutional neural networks (CNN) of two types: GoogLeNet and AlexNet. Due to the very large size of images of cytological specimen (on average 200000 × 100000 pixels), they were divided into smaller patches of size 256 × 256 pixels. Breast cancer classification usually is based on morphometric features of nuclei. Therefore, training and validation patches were selected using Support Vector Machine (SVM) so that suitable amount of cell material was depicted. Neural classifiers were tuned using GPU accelerated implementation of gradient descent algorithm. Training error was defined as a cross-entropy classification loss. Classification accuracy was defined as the percentage ratio of successfully classified validation patches to the total number of validation patches. The best accuracy rate of 83% was obtained by GoogLeNet model. We observed that more misclassified patches belong to malignant cases.

  19. Very Deep Convolutional Neural Networks for Morphologic Classification of Erythrocytes.

    Science.gov (United States)

    Durant, Thomas J S; Olson, Eben M; Schulz, Wade L; Torres, Richard

    2017-12-01

    Morphologic profiling of the erythrocyte population is a widely used and clinically valuable diagnostic modality, but one that relies on a slow manual process associated with significant labor cost and limited reproducibility. Automated profiling of erythrocytes from digital images by capable machine learning approaches would augment the throughput and value of morphologic analysis. To this end, we sought to evaluate the performance of leading implementation strategies for convolutional neural networks (CNNs) when applied to classification of erythrocytes based on morphology. Erythrocytes were manually classified into 1 of 10 classes using a custom-developed Web application. Using recent literature to guide architectural considerations for neural network design, we implemented a "very deep" CNN, consisting of >150 layers, with dense shortcut connections. The final database comprised 3737 labeled cells. Ensemble model predictions on unseen data demonstrated a harmonic mean of recall and precision metrics of 92.70% and 89.39%, respectively. Of the 748 cells in the test set, 23 misclassification errors were made, with a correct classification frequency of 90.60%, represented as a harmonic mean across the 10 morphologic classes. These findings indicate that erythrocyte morphology profiles could be measured with a high degree of accuracy with "very deep" CNNs. Further, these data support future efforts to expand classes and optimize practical performance in a clinical environment as a prelude to full implementation as a clinical tool. © 2017 American Association for Clinical Chemistry.

  20. Protein Secondary Structure Prediction Using Deep Convolutional Neural Fields.

    Science.gov (United States)

    Wang, Sheng; Peng, Jian; Ma, Jianzhu; Xu, Jinbo

    2016-01-11

    Protein secondary structure (SS) prediction is important for studying protein structure and function. When only the sequence (profile) information is used as input feature, currently the best predictors can obtain ~80% Q3 accuracy, which has not been improved in the past decade. Here we present DeepCNF (Deep Convolutional Neural Fields) for protein SS prediction. DeepCNF is a Deep Learning extension of Conditional Neural Fields (CNF), which is an integration of Conditional Random Fields (CRF) and shallow neural networks. DeepCNF can model not only complex sequence-structure relationship by a deep hierarchical architecture, but also interdependency between adjacent SS labels, so it is much more powerful than CNF. Experimental results show that DeepCNF can obtain ~84% Q3 accuracy, ~85% SOV score, and ~72% Q8 accuracy, respectively, on the CASP and CAMEO test proteins, greatly outperforming currently popular predictors. As a general framework, DeepCNF can be used to predict other protein structure properties such as contact number, disorder regions, and solvent accessibility.

  1. Classifying magnetic resonance image modalities with convolutional neural networks

    Science.gov (United States)

    Remedios, Samuel; Pham, Dzung L.; Butman, John A.; Roy, Snehashis

    2018-02-01

    Magnetic Resonance (MR) imaging allows the acquisition of images with different contrast properties depending on the acquisition protocol and the magnetic properties of tissues. Many MR brain image processing techniques, such as tissue segmentation, require multiple MR contrasts as inputs, and each contrast is treated differently. Thus it is advantageous to automate the identification of image contrasts for various purposes, such as facilitating image processing pipelines, and managing and maintaining large databases via content-based image retrieval (CBIR). Most automated CBIR techniques focus on a two-step process: extracting features from data and classifying the image based on these features. We present a novel 3D deep convolutional neural network (CNN)- based method for MR image contrast classification. The proposed CNN automatically identifies the MR contrast of an input brain image volume. Specifically, we explored three classification problems: (1) identify T1-weighted (T1-w), T2-weighted (T2-w), and fluid-attenuated inversion recovery (FLAIR) contrasts, (2) identify pre vs postcontrast T1, (3) identify pre vs post-contrast FLAIR. A total of 3418 image volumes acquired from multiple sites and multiple scanners were used. To evaluate each task, the proposed model was trained on 2137 images and tested on the remaining 1281 images. Results showed that image volumes were correctly classified with 97.57% accuracy.

  2. Mobile Stride Length Estimation With Deep Convolutional Neural Networks.

    Science.gov (United States)

    Hannink, Julius; Kautz, Thomas; Pasluosta, Cristian F; Barth, Jens; Schulein, Samuel; GaBmann, Karl-Gunter; Klucken, Jochen; Eskofier, Bjoern M

    2018-03-01

    Accurate estimation of spatial gait characteristics is critical to assess motor impairments resulting from neurological or musculoskeletal disease. Currently, however, methodological constraints limit clinical applicability of state-of-the-art double integration approaches to gait patterns with a clear zero-velocity phase. We describe a novel approach to stride length estimation that uses deep convolutional neural networks to map stride-specific inertial sensor data to the resulting stride length. The model is trained on a publicly available and clinically relevant benchmark dataset consisting of 1220 strides from 101 geriatric patients. Evaluation is done in a tenfold cross validation and for three different stride definitions. Even though best results are achieved with strides defined from midstance to midstance with average accuracy and precision of , performance does not strongly depend on stride definition. The achieved precision outperforms state-of-the-art methods evaluated on the same benchmark dataset by . Due to the independence of stride definition, the proposed method is not subject to the methodological constrains that limit applicability of state-of-the-art double integration methods. Furthermore, it was possible to improve precision on the benchmark dataset. With more precise mobile stride length estimation, new insights to the progression of neurological disease or early indications might be gained. Due to the independence of stride definition, previously uncharted diseases in terms of mobile gait analysis can now be investigated by retraining and applying the proposed method.

  3. Low-complexity object detection with deep convolutional neural network for embedded systems

    Science.gov (United States)

    Tripathi, Subarna; Kang, Byeongkeun; Dane, Gokce; Nguyen, Truong

    2017-09-01

    We investigate low-complexity convolutional neural networks (CNNs) for object detection for embedded vision applications. It is well-known that consolidation of an embedded system for CNN-based object detection is more challenging due to computation and memory requirement comparing with problems like image classification. To achieve these requirements, we design and develop an end-to-end TensorFlow (TF)-based fully-convolutional deep neural network for generic object detection task inspired by one of the fastest framework, YOLO.1 The proposed network predicts the localization of every object by regressing the coordinates of the corresponding bounding box as in YOLO. Hence, the network is able to detect any objects without any limitations in the size of the objects. However, unlike YOLO, all the layers in the proposed network is fully-convolutional. Thus, it is able to take input images of any size. We pick face detection as an use case. We evaluate the proposed model for face detection on FDDB dataset and Widerface dataset. As another use case of generic object detection, we evaluate its performance on PASCAL VOC dataset. The experimental results demonstrate that the proposed network can predict object instances of different sizes and poses in a single frame. Moreover, the results show that the proposed method achieves comparative accuracy comparing with the state-of-the-art CNN-based object detection methods while reducing the model size by 3× and memory-BW by 3 - 4× comparing with one of the best real-time CNN-based object detectors, YOLO. Our 8-bit fixed-point TF-model provides additional 4× memory reduction while keeping the accuracy nearly as good as the floating-point model. Moreover, the fixed- point model is capable of achieving 20× faster inference speed comparing with the floating-point model. Thus, the proposed method is promising for embedded implementations.

  4. Invariant moments based convolutional neural networks for image analysis

    Directory of Open Access Journals (Sweden)

    Vijayalakshmi G.V. Mahesh

    2017-01-01

    Full Text Available The paper proposes a method using convolutional neural network to effectively evaluate the discrimination between face and non face patterns, gender classification using facial images and facial expression recognition. The novelty of the method lies in the utilization of the initial trainable convolution kernels coefficients derived from the zernike moments by varying the moment order. The performance of the proposed method was compared with the convolutional neural network architecture that used random kernels as initial training parameters. The multilevel configuration of zernike moments was significant in extracting the shape information suitable for hierarchical feature learning to carry out image analysis and classification. Furthermore the results showed an outstanding performance of zernike moment based kernels in terms of the computation time and classification accuracy.

  5. Single image super-resolution based on convolutional neural networks

    Science.gov (United States)

    Zou, Lamei; Luo, Ming; Yang, Weidong; Li, Peng; Jin, Liujia

    2018-03-01

    We present a deep learning method for single image super-resolution (SISR). The proposed approach learns end-to-end mapping between low-resolution (LR) images and high-resolution (HR) images. The mapping is represented as a deep convolutional neural network which inputs the LR image and outputs the HR image. Our network uses 5 convolution layers, which kernels size include 5×5, 3×3 and 1×1. In our proposed network, we use residual-learning and combine different sizes of convolution kernels at the same layer. The experiment results show that our proposed method performs better than the existing methods in reconstructing quality index and human visual effects on benchmarked images.

  6. Change detection in multitemporal synthetic aperture radar images using dual-channel convolutional neural network

    Science.gov (United States)

    Liu, Tao; Li, Ying; Cao, Ying; Shen, Qiang

    2017-10-01

    This paper proposes a model of dual-channel convolutional neural network (CNN) that is designed for change detection in SAR images, in an effort to acquire higher detection accuracy and lower misclassification rate. This network model contains two parallel CNN channels, which can extract deep features from two multitemporal SAR images. For comparison and validation, the proposed method is tested along with other change detection algorithms on both simulated SAR images and real-world SAR images captured by different sensors. The experimental results demonstrate that the presented method outperforms the state-of-the-art techniques by a considerable margin.

  7. Spacings and pair correlations for finite Bernoulli convolutions

    International Nuclear Information System (INIS)

    Benjamini, Itai; Solomyak, Boris

    2009-01-01

    We consider finite Bernoulli convolutions with a parameter 1/2 N . These sequences are uniformly distributed with respect to the infinite Bernoulli convolution measure ν λ , as N → ∞. Numerical evidence suggests that for a generic λ, the distribution of spacings between appropriately rescaled points is Poissonian. We obtain some partial results in this direction; for instance, we show that, on average, the pair correlations do not exhibit attraction or repulsion in the limit. On the other hand, for certain algebraic λ the behaviour is totally different

  8. A New Reverberator Based on Variable Sparsity Convolution

    DEFF Research Database (Denmark)

    Holm-Rasmussen, Bo; Lehtonen, Heidi-Maria; Välimäki, Vesa

    2013-01-01

    FIR filter coefficients are selected from a velvet noise sequence, which consists of ones, minus ones, and zeros only. In this application, it is sufficient perceptually to use very sparse velvet noise sequences having only about 0.1 to 0.2% non-zero elements, with increasing sparsity along...... the impulse response. The algorithm yields a parametric approximation of the late part of the impulse response, which is more than 100 times more efficient computationally than the direct convolution. The computational load of the proposed algorithm is comparable to that of FFT-based partitioned convolution...

  9. Convolutional cylinder-type block-circulant cycle codes

    Directory of Open Access Journals (Sweden)

    Mohammad Gholami

    2013-06-01

    Full Text Available In this paper, we consider a class of column-weight two quasi-cyclic low-density paritycheck codes in which the girth can be large enough, as an arbitrary multiple of 8. Then we devote a convolutional form to these codes, such that their generator matrix can be obtained by elementary row and column operations on the parity-check matrix. Finally, we show that the free distance of the convolutional codes is equal to the minimum distance of their block counterparts.

  10. Weed Growth Stage Estimator Using Deep Convolutional Neural Networks

    DEFF Research Database (Denmark)

    Teimouri, Nima; Dyrmann, Mads; Nielsen, Per Rydahl

    2018-01-01

    conditions with regards to soil types, resolution and light settings. Then, 9649 of these images were used for training the computer, which automatically divided the weeds into nine growth classes. The performance of this proposed convolutional neural network approach was evaluated on a further set of 2516...... in estimating the number of leaves and 96% accuracy when accepting a deviation of two leaves. These results show that this new method of using deep convolutional neural networks has a relatively high ability to estimate early growth stages across a wide variety of weed species....

  11. Deep Convolutional Neural Networks: Structure, Feature Extraction and Training

    Directory of Open Access Journals (Sweden)

    Namatēvs Ivars

    2017-12-01

    Full Text Available Deep convolutional neural networks (CNNs are aimed at processing data that have a known network like topology. They are widely used to recognise objects in images and diagnose patterns in time series data as well as in sensor data classification. The aim of the paper is to present theoretical and practical aspects of deep CNNs in terms of convolution operation, typical layers and basic methods to be used for training and learning. Some practical applications are included for signal and image classification. Finally, the present paper describes the proposed block structure of CNN for classifying crucial features from 3D sensor data.

  12. Very deep recurrent convolutional neural network for object recognition

    Science.gov (United States)

    Brahimi, Sourour; Ben Aoun, Najib; Ben Amar, Chokri

    2017-03-01

    In recent years, Computer vision has become a very active field. This field includes methods for processing, analyzing, and understanding images. The most challenging problems in computer vision are image classification and object recognition. This paper presents a new approach for object recognition task. This approach exploits the success of the Very Deep Convolutional Neural Network for object recognition. In fact, it improves the convolutional layers by adding recurrent connections. This proposed approach was evaluated on two object recognition benchmarks: Pascal VOC 2007 and CIFAR-10. The experimental results prove the efficiency of our method in comparison with the state of the art methods.

  13. Spectral interpolation - Zero fill or convolution. [image processing

    Science.gov (United States)

    Forman, M. L.

    1977-01-01

    Zero fill, or augmentation by zeros, is a method used in conjunction with fast Fourier transforms to obtain spectral spacing at intervals closer than obtainable from the original input data set. In the present paper, an interpolation technique (interpolation by repetitive convolution) is proposed which yields values accurate enough for plotting purposes and which lie within the limits of calibration accuracies. The technique is shown to operate faster than zero fill, since fewer operations are required. The major advantages of interpolation by repetitive convolution are that efficient use of memory is possible (thus avoiding the difficulties encountered in decimation in time FFTs) and that is is easy to implement.

  14. Application of the Convolution Formalism to the Ocean Tide Potential: Results from the Gravity and Recovery and Climate Experiment (GRACE)

    Science.gov (United States)

    Desai, S. D.; Yuan, D. -N.

    2006-01-01

    A computationally efficient approach to reducing omission errors in ocean tide potential models is derived and evaluated using data from the Gravity Recovery and Climate Experiment (GRACE) mission. Ocean tide height models are usually explicitly available at a few frequencies, and a smooth unit response is assumed to infer the response across the tidal spectrum. The convolution formalism of Munk and Cartwright (1966) models this response function with a Fourier series. This allows the total ocean tide height, and therefore the total ocean tide potential, to be modeled as a weighted sum of past, present, and future values of the tide-generating potential. Previous applications of the convolution formalism have usually been limited to tide height models, but we extend it to ocean tide potential models. We use luni-solar ephemerides to derive the required tide-generating potential so that the complete spectrum of the ocean tide potential is efficiently represented. In contrast, the traditionally adopted harmonic model of the ocean tide potential requires the explicit sum of the contributions from individual tidal frequencies. It is therefore subject to omission errors from neglected frequencies and is computationally more intensive. Intersatellite range rate data from the GRACE mission are used to compare convolution and harmonic models of the ocean tide potential. The monthly range rate residual variance is smaller by 4-5%, and the daily residual variance is smaller by as much as 15% when using the convolution model than when using a harmonic model that is defined by twice the number of parameters.

  15. Efficient forward propagation of time-sequences in convolutional neural networks using Deep Shifting

    NARCIS (Netherlands)

    K.L. Groenland (Koen); S.M. Bohte (Sander)

    2016-01-01

    textabstractWhen a Convolutional Neural Network is used for on-the-fly evaluation of continuously updating time-sequences, many redundant convolution operations are performed. We propose the method of Deep Shifting, which remembers previously calculated results of convolution operations in order

  16. Traffic Command Gesture Recognition for Virtual Urban Scenes Based on a Spatiotemporal Convolution Neural Network

    Directory of Open Access Journals (Sweden)

    Chunyong Ma

    2018-01-01

    Full Text Available Intelligent recognition of traffic police command gestures increases authenticity and interactivity in virtual urban scenes. To actualize real-time traffic gesture recognition, a novel spatiotemporal convolution neural network (ST-CNN model is presented. We utilized Kinect 2.0 to construct a traffic police command gesture skeleton (TPCGS dataset collected from 10 volunteers. Subsequently, convolution operations on the locational change of each skeletal point were performed to extract temporal features, analyze the relative positions of skeletal points, and extract spatial features. After temporal and spatial features based on the three-dimensional positional information of traffic police skeleton points were extracted, the ST-CNN model classified positional information into eight types of Chinese traffic police gestures. The test accuracy of the ST-CNN model was 96.67%. In addition, a virtual urban traffic scene in which real-time command tests were carried out was set up, and a real-time test accuracy rate of 93.0% was achieved. The proposed ST-CNN model ensured a high level of accuracy and robustness. The ST-CNN model recognized traffic command gestures, and such recognition was found to control vehicles in virtual traffic environments, which enriches the interactive mode of the virtual city scene. Traffic command gesture recognition contributes to smart city construction.

  17. Discrete singular convolution for the generalized variable-coefficient ...

    African Journals Online (AJOL)

    Numerical solutions of the generalized variable-coefficient Korteweg-de Vries equation are obtained using a discrete singular convolution and a fourth order singly diagonally implicit Runge-Kutta method for space and time discretisation, respectively. The theoretical convergence of the proposed method is rigorously ...

  18. Face recognition: a convolutional neural-network approach.

    Science.gov (United States)

    Lawrence, S; Giles, C L; Tsoi, A C; Back, A D

    1997-01-01

    We present a hybrid neural-network for human face recognition which compares favourably with other methods. The system combines local image sampling, a self-organizing map (SOM) neural network, and a convolutional neural network. The SOM provides a quantization of the image samples into a topological space where inputs that are nearby in the original space are also nearby in the output space, thereby providing dimensionality reduction and invariance to minor changes in the image sample, and the convolutional neural network provides partial invariance to translation, rotation, scale, and deformation. The convolutional network extracts successively larger features in a hierarchical set of layers. We present results using the Karhunen-Loeve transform in place of the SOM, and a multilayer perceptron (MLP) in place of the convolutional network for comparison. We use a database of 400 images of 40 individuals which contains quite a high degree of variability in expression, pose, and facial details. We analyze the computational complexity and discuss how new classes could be added to the trained recognizer.

  19. Training Convolutional Neural Networks for Translational Invariance on SAR ATR

    DEFF Research Database (Denmark)

    Malmgren-Hansen, David; Engholm, Rasmus; Østergaard Pedersen, Morten

    2016-01-01

    In this paper we present a comparison of the robustness of Convolutional Neural Networks (CNN) to other classifiers in the presence of uncertainty of the objects localization in SAR image. We present a framework for simulating simple SAR images, translating the object of interest systematically...

  20. An Interactive Graphics Program for Assistance in Learning Convolution.

    Science.gov (United States)

    Frederick, Dean K.; Waag, Gary L.

    1980-01-01

    A program has been written for the interactive computer graphics facility at Rensselaer Polytechnic Institute that is designed to assist the user in learning the mathematical technique of convolving two functions. Because convolution can be represented graphically by a sequence of steps involving folding, shifting, multiplying, and integration, it…

  1. Diffraction and Dirchlet problem for parameter-elliptic convolution ...

    African Journals Online (AJOL)

    In this paper we evaluate the difference between the inverse operators of a Dirichlet problem and of a diffraction problem for parameter-elliptic convolution operators with constant symbols. We prove that the inverse operator of a Dirichlet problem can be obtained as a limit case of such a diffraction problem. Quaestiones ...

  2. Deep convolutional neural networks for detection of rail surface defects

    NARCIS (Netherlands)

    Faghih Roohi, S.; Hajizadeh, S.; Nunez Vicencio, Alfredo; Babuska, R.; De Schutter, B.H.K.; Estevez, Pablo A.; Angelov, Plamen P.; Del Moral Hernandez, Emilio

    2016-01-01

    In this paper, we propose a deep convolutional neural network solution to the analysis of image data for the detection of rail surface defects. The images are obtained from many hours of automated video recordings. This huge amount of data makes it impossible to manually inspect the images and

  3. Symbol Stream Combining in a Convolutionally Coded System

    Science.gov (United States)

    Mceliece, R. J.; Pollara, F.; Swanson, L.

    1985-01-01

    Symbol stream combining has been proposed as a method for arraying signals received at different antennas. If convolutional coding and Viterbi decoding are used, it is shown that a Viterbi decoder based on the proposed weighted sum of symbol streams yields maximum likelihood decisions.

  4. Two-level convolution formula for nuclear structure function

    Science.gov (United States)

    Ma, Boqiang

    1990-05-01

    A two-level convolution formula for the nuclear structure function is derived in considering the nucleus as a composite system of baryon-mesons which are also composite systems of quark-gluons again. The results show that the European Muon Colaboration effect can not be explained by the nuclear effects as nucleon Fermi motion and nuclear binding contributions.

  5. Two-level convolution formula for nuclear structure function

    International Nuclear Information System (INIS)

    Ma Boqiang

    1990-01-01

    A two-level convolution formula for the nuclear structure function is derived in considering the nucleus as a composite system of baryon-mesons which are also composite systems of quark-gluons again. The results show that the European Muon Colaboration effect can not be explained by the nuclear effects as nucleon Fermi motion and nuclear binding contributions

  6. Plant species classification using deep convolutional neural network

    DEFF Research Database (Denmark)

    Dyrmann, Mads; Karstoft, Henrik; Midtiby, Henrik Skov

    2016-01-01

    Information on which weed species are present within agricultural fields is important for site specific weed management. This paper presents a method that is capable of recognising plant species in colour images by using a convolutional neural network. The network is built from scratch trained an...

  7. Pattern Recognition of Momentary Mental Workload Based on Multi-Channel Electrophysiological Data and Ensemble Convolutional Neural Networks.

    Science.gov (United States)

    Zhang, Jianhua; Li, Sunan; Wang, Rubin

    2017-01-01

    In this paper, we deal with the Mental Workload (MWL) classification problem based on the measured physiological data. First we discussed the optimal depth (i.e., the number of hidden layers) and parameter optimization algorithms for the Convolutional Neural Networks (CNN). The base CNNs designed were tested according to five classification performance indices, namely Accuracy, Precision, F-measure, G-mean, and required training time. Then we developed an Ensemble Convolutional Neural Network (ECNN) to enhance the accuracy and robustness of the individual CNN model. For the ECNN design, three model aggregation approaches (weighted averaging, majority voting and stacking) were examined and a resampling strategy was used to enhance the diversity of individual CNN models. The results of MWL classification performance comparison indicated that the proposed ECNN framework can effectively improve MWL classification performance and is featured by entirely automatic feature extraction and MWL classification, when compared with traditional machine learning methods.

  8. Extreme-value limit of the convolution of exponential and multivariate normal distributions: Link to the Hüsler–Reiß distribution

    KAUST Repository

    Krupskii, Pavel; Joe, Harry; Lee, David; Genton, Marc G.

    2017-01-01

    The multivariate Hüsler–Reiß copula is obtained as a direct extreme-value limit from the convolution of a multivariate normal random vector and an exponential random variable multiplied by a vector of constants. It is shown how the set of Hüsler–Reiß parameters can be mapped to the parameters of this convolution model. Assuming there are no singular components in the Hüsler–Reiß copula, the convolution model leads to exact and approximate simulation methods. An application of simulation is to check if the Hüsler–Reiß copula with different parsimonious dependence structures provides adequate fit to some data consisting of multivariate extremes.

  9. Extreme-value limit of the convolution of exponential and multivariate normal distributions: Link to the Hüsler–Reiß distribution

    KAUST Repository

    Krupskii, Pavel

    2017-11-02

    The multivariate Hüsler–Reiß copula is obtained as a direct extreme-value limit from the convolution of a multivariate normal random vector and an exponential random variable multiplied by a vector of constants. It is shown how the set of Hüsler–Reiß parameters can be mapped to the parameters of this convolution model. Assuming there are no singular components in the Hüsler–Reiß copula, the convolution model leads to exact and approximate simulation methods. An application of simulation is to check if the Hüsler–Reiß copula with different parsimonious dependence structures provides adequate fit to some data consisting of multivariate extremes.

  10. Multiscale Rotation-Invariant Convolutional Neural Networks for Lung Texture Classification.

    Science.gov (United States)

    Wang, Qiangchang; Zheng, Yuanjie; Yang, Gongping; Jin, Weidong; Chen, Xinjian; Yin, Yilong

    2018-01-01

    We propose a new multiscale rotation-invariant convolutional neural network (MRCNN) model for classifying various lung tissue types on high-resolution computed tomography. MRCNN employs Gabor-local binary pattern that introduces a good property in image analysis-invariance to image scales and rotations. In addition, we offer an approach to deal with the problems caused by imbalanced number of samples between different classes in most of the existing works, accomplished by changing the overlapping size between the adjacent patches. Experimental results on a public interstitial lung disease database show a superior performance of the proposed method to state of the art.

  11. Convolutional Deep Belief Networks for Single-Cell/Object Tracking in Computational Biology and Computer Vision.

    Science.gov (United States)

    Zhong, Bineng; Pan, Shengnan; Zhang, Hongbo; Wang, Tian; Du, Jixiang; Chen, Duansheng; Cao, Liujuan

    2016-01-01

    In this paper, we propose deep architecture to dynamically learn the most discriminative features from data for both single-cell and object tracking in computational biology and computer vision. Firstly, the discriminative features are automatically learned via a convolutional deep belief network (CDBN). Secondly, we design a simple yet effective method to transfer features learned from CDBNs on the source tasks for generic purpose to the object tracking tasks using only limited amount of training data. Finally, to alleviate the tracker drifting problem caused by model updating, we jointly consider three different types of positive samples. Extensive experiments validate the robustness and effectiveness of the proposed method.

  12. Deep convolutional neural networks for estimating porous material parameters with ultrasound tomography

    Science.gov (United States)

    Lähivaara, Timo; Kärkkäinen, Leo; Huttunen, Janne M. J.; Hesthaven, Jan S.

    2018-02-01

    We study the feasibility of data based machine learning applied to ultrasound tomography to estimate water-saturated porous material parameters. In this work, the data to train the neural networks is simulated by solving wave propagation in coupled poroviscoelastic-viscoelastic-acoustic media. As the forward model, we consider a high-order discontinuous Galerkin method while deep convolutional neural networks are used to solve the parameter estimation problem. In the numerical experiment, we estimate the material porosity and tortuosity while the remaining parameters which are of less interest are successfully marginalized in the neural networks-based inversion. Computational examples confirms the feasibility and accuracy of this approach.

  13. Automatic Seismic-Event Classification with Convolutional Neural Networks.

    Science.gov (United States)

    Bueno Rodriguez, A.; Titos Luzón, M.; Garcia Martinez, L.; Benitez, C.; Ibáñez, J. M.

    2017-12-01

    Active volcanoes exhibit a wide range of seismic signals, providing vast amounts of unlabelled volcano-seismic data that can be analyzed through the lens of artificial intelligence. However, obtaining high-quality labelled data is time-consuming and expensive. Deep neural networks can process data in their raw form, compute high-level features and provide a better representation of the input data distribution. These systems can be deployed to classify seismic data at scale, enhance current early-warning systems and build extensive seismic catalogs. In this research, we aim to classify spectrograms from seven different seismic events registered at "Volcán de Fuego" (Colima, Mexico), during four eruptive periods. Our approach is based on convolutional neural networks (CNNs), a sub-type of deep neural networks that can exploit grid structure from the data. Volcano-seismic signals can be mapped into a grid-like structure using the spectrogram: a representation of the temporal evolution in terms of time and frequency. Spectrograms were computed from the data using Hamming windows with 4 seconds length, 2.5 seconds overlapping and 128 points FFT resolution. Results are compared to deep neural networks, random forest and SVMs. Experiments show that CNNs can exploit temporal and frequency information, attaining a classification accuracy of 93%, similar to deep networks 91% but outperforming SVM and random forest. These results empirically show that CNNs are powerful models to classify a wide range of volcano-seismic signals, and achieve good generalization. Furthermore, volcano-seismic spectrograms contains useful discriminative information for the CNN, as higher layers of the network combine high-level features computed for each frequency band, helping to detect simultaneous events in time. Being at the intersection of deep learning and geophysics, this research enables future studies of how CNNs can be used in volcano monitoring to accurately determine the detection and

  14. Psoriasis skin biopsy image segmentation using Deep Convolutional Neural Network.

    Science.gov (United States)

    Pal, Anabik; Garain, Utpal; Chandra, Aditi; Chatterjee, Raghunath; Senapati, Swapan

    2018-06-01

    Development of machine assisted tools for automatic analysis of psoriasis skin biopsy image plays an important role in clinical assistance. Development of automatic approach for accurate segmentation of psoriasis skin biopsy image is the initial prerequisite for developing such system. However, the complex cellular structure, presence of imaging artifacts, uneven staining variation make the task challenging. This paper presents a pioneering attempt for automatic segmentation of psoriasis skin biopsy images. Several deep neural architectures are tried for segmenting psoriasis skin biopsy images. Deep models are used for classifying the super-pixels generated by Simple Linear Iterative Clustering (SLIC) and the segmentation performance of these architectures is compared with the traditional hand-crafted feature based classifiers built on popularly used classifiers like K-Nearest Neighbor (KNN), Support Vector Machine (SVM) and Random Forest (RF). A U-shaped Fully Convolutional Neural Network (FCN) is also used in an end to end learning fashion where input is the original color image and the output is the segmentation class map for the skin layers. An annotated real psoriasis skin biopsy image data set of ninety (90) images is developed and used for this research. The segmentation performance is evaluated with two metrics namely, Jaccard's Coefficient (JC) and the Ratio of Correct Pixel Classification (RCPC) accuracy. The experimental results show that the CNN based approaches outperform the traditional hand-crafted feature based classification approaches. The present research shows that practical system can be developed for machine assisted analysis of psoriasis disease. Copyright © 2018 Elsevier B.V. All rights reserved.

  15. Deep convolutional neural networks for annotating gene expression patterns in the mouse brain.

    Science.gov (United States)

    Zeng, Tao; Li, Rongjian; Mukkamala, Ravi; Ye, Jieping; Ji, Shuiwang

    2015-05-07

    Profiling gene expression in brain structures at various spatial and temporal scales is essential to understanding how genes regulate the development of brain structures. The Allen Developing Mouse Brain Atlas provides high-resolution 3-D in situ hybridization (ISH) gene expression patterns in multiple developing stages of the mouse brain. Currently, the ISH images are annotated with anatomical terms manually. In this paper, we propose a computational approach to annotate gene expression pattern images in the mouse brain at various structural levels over the course of development. We applied deep convolutional neural network that was trained on a large set of natural images to extract features from the ISH images of developing mouse brain. As a baseline representation, we applied invariant image feature descriptors to capture local statistics from ISH images and used the bag-of-words approach to build image-level representations. Both types of features from multiple ISH image sections of the entire brain were then combined to build 3-D, brain-wide gene expression representations. We employed regularized learning methods for discriminating gene expression patterns in different brain structures. Results show that our approach of using convolutional model as feature extractors achieved superior performance in annotating gene expression patterns at multiple levels of brain structures throughout four developing ages. Overall, we achieved average AUC of 0.894 ± 0.014, as compared with 0.820 ± 0.046 yielded by the bag-of-words approach. Deep convolutional neural network model trained on natural image sets and applied to gene expression pattern annotation tasks yielded superior performance, demonstrating its transfer learning property is applicable to such biological image sets.

  16. Convolutional Sparse Coding for Static and Dynamic Images Analysis

    Directory of Open Access Journals (Sweden)

    B. A. Knyazev

    2014-01-01

    Full Text Available The objective of this work is to improve performance of static and dynamic objects recognition. For this purpose a new image representation model and a transformation algorithm are proposed. It is examined and illustrated that limitations of previous methods make it difficult to achieve this objective. Static images, specifically handwritten digits of the widely used MNIST dataset, are the primary focus of this work. Nevertheless, preliminary qualitative results of image sequences analysis based on the suggested model are presented.A general analytical form of the Gabor function, often employed to generate filters, is described and discussed. In this research, this description is required for computing parameters of responses returned by our algorithm. The recursive convolution operator is introduced, which allows extracting free shape features of visual objects. The developed parametric representation model is compared with sparse coding based on energy function minimization.In the experimental part of this work, errors of estimating the parameters of responses are determined. Also, parameters statistics and their correlation coefficients for more than 106 responses extracted from the MNIST dataset are calculated. It is demonstrated that these data correspond well with previous research studies on Gabor filters as well as with works on visual cortex primary cells of mammals, in which similar responses were observed. A comparative test of the developed model with three other approaches is conducted; speed and accuracy scores of handwritten digits classification are presented. A support vector machine with a linear or radial basic function is used for classification of images and their representations while principal component analysis is used in some cases to prepare data beforehand. High accuracy is not attained due to the specific difficulties of combining our model with a support vector machine (a 3.99% error rate. However, another method is

  17. An Implementation of Error Minimization Data Transmission in OFDM using Modified Convolutional Code

    Directory of Open Access Journals (Sweden)

    Hendy Briantoro

    2016-04-01

    Full Text Available This paper presents about error minimization in OFDM system. In conventional system, usually using channel coding such as BCH Code or Convolutional Code. But, performance BCH Code or Convolutional Code is not good in implementation of OFDM System. Error bits of OFDM system without channel coding is 5.77%. Then, we used convolutional code with code rate 1/2, it can reduce error bitsonly up to 3.85%. So, we proposed OFDM system with Modified Convolutional Code. In this implementation, we used Software Define Radio (SDR, namely Universal Software Radio Peripheral (USRP NI 2920 as the transmitter and receiver. The result of OFDM system using Modified Convolutional Code with code rate is able recover all character received so can decrease until 0% error bit. Increasing performance of Modified Convolutional Code is about 1 dB in BER of 10-4 from BCH Code and Convolutional Code. So, performance of Modified Convolutional better than BCH Code or Convolutional Code. Keywords: OFDM, BCH Code, Convolutional Code, Modified Convolutional Code, SDR, USRP

  18. Mathematical models of decision making on space vehicle motion control at fuzzy conditions

    International Nuclear Information System (INIS)

    Arslanov, M.Z.; Ismail, E.E.; Oryspaev, D.O.

    2005-01-01

    the structure of decision making for control of spacecraft motion is considered. Mathematical models of decision making problems are investigated. New methods of criteria convolution are received. New convolution have properties of smoothness. (author)

  19. Voltage measurements at the vacuum post-hole convolute of the Z pulsed-power accelerator

    Directory of Open Access Journals (Sweden)

    E. M. Waisman

    2014-12-01

    Full Text Available Presented are voltage measurements taken near the load region on the Z pulsed-power accelerator using an inductive voltage monitor (IVM. Specifically, the IVM was connected to, and thus monitored the voltage at, the bottom level of the accelerator’s vacuum double post-hole convolute. Additional voltage and current measurements were taken at the accelerator’s vacuum-insulator stack (at a radius of 1.6 m by using standard D-dot and B-dot probes, respectively. During postprocessing, the measurements taken at the stack were translated to the location of the IVM measurements by using a lossless propagation model of the Z accelerator’s magnetically insulated transmission lines (MITLs and a lumped inductor model of the vacuum post-hole convolute. Across a wide variety of experiments conducted on the Z accelerator, the voltage histories obtained from the IVM and the lossless propagation technique agree well in overall shape and magnitude. However, large-amplitude, high-frequency oscillations are more pronounced in the IVM records. It is unclear whether these larger oscillations represent true voltage oscillations at the convolute or if they are due to noise pickup and/or transit-time effects and other resonant modes in the IVM. Results using a transit-time-correction technique and Fourier analysis support the latter. Regardless of which interpretation is correct, both true voltage oscillations and the excitement of resonant modes could be the result of transient electrical breakdowns in the post-hole convolute, though more information is required to determine definitively if such breakdowns occurred. Despite the larger oscillations in the IVM records, the general agreement found between the lossless propagation results and the results of the IVM shows that large voltages are transmitted efficiently through the MITLs on Z. These results are complementary to previous studies [R. D. McBride et al., Phys. Rev. ST Accel. Beams 13, 120401 (2010

  20. A Data-Driven Response Virtual Sensor Technique with Partial Vibration Measurements Using Convolutional Neural Network

    Science.gov (United States)

    Sun, Shan-Bin; He, Yuan-Yuan; Zhou, Si-Da; Yue, Zhen-Jiang

    2017-01-01

    Measurement of dynamic responses plays an important role in structural health monitoring, damage detection and other fields of research. However, in aerospace engineering, the physical sensors are limited in the operational conditions of spacecraft, due to the severe environment in outer space. This paper proposes a virtual sensor model with partial vibration measurements using a convolutional neural network. The transmissibility function is employed as prior knowledge. A four-layer neural network with two convolutional layers, one fully connected layer, and an output layer is proposed as the predicting model. Numerical examples of two different structural dynamic systems demonstrate the performance of the proposed approach. The excellence of the novel technique is further indicated using a simply supported beam experiment comparing to a modal-model-based virtual sensor, which uses modal parameters, such as mode shapes, for estimating the responses of the faulty sensors. The results show that the presented data-driven response virtual sensor technique can predict structural response with high accuracy. PMID:29231868

  1. Using Convolutional Neural Network Filters to Measure Left-Right Mirror Symmetry in Images

    Directory of Open Access Journals (Sweden)

    Anselm Brachmann

    2016-12-01

    Full Text Available We propose a method for measuring symmetry in images by using filter responses from Convolutional Neural Networks (CNNs. The aim of the method is to model human perception of left/right symmetry as closely as possible. Using the Convolutional Neural Network (CNN approach has two main advantages: First, CNN filter responses closely match the responses of neurons in the human visual system; they take information on color, edges and texture into account simultaneously. Second, we can measure higher-order symmetry, which relies not only on color, edges and texture, but also on the shapes and objects that are depicted in images. We validated our algorithm on a dataset of 300 music album covers, which were rated according to their symmetry by 20 human observers, and compared results with those from a previously proposed method. With our method, human perception of symmetry can be predicted with high accuracy. Moreover, we demonstrate that the inclusion of features from higher CNN layers, which encode more abstract image content, increases the performance further. In conclusion, we introduce a model of left/right symmetry that closely models human perception of symmetry in CD album covers.

  2. A Data-Driven Response Virtual Sensor Technique with Partial Vibration Measurements Using Convolutional Neural Network.

    Science.gov (United States)

    Sun, Shan-Bin; He, Yuan-Yuan; Zhou, Si-Da; Yue, Zhen-Jiang

    2017-12-12

    Measurement of dynamic responses plays an important role in structural health monitoring, damage detection and other fields of research. However, in aerospace engineering, the physical sensors are limited in the operational conditions of spacecraft, due to the severe environment in outer space. This paper proposes a virtual sensor model with partial vibration measurements using a convolutional neural network. The transmissibility function is employed as prior knowledge. A four-layer neural network with two convolutional layers, one fully connected layer, and an output layer is proposed as the predicting model. Numerical examples of two different structural dynamic systems demonstrate the performance of the proposed approach. The excellence of the novel technique is further indicated using a simply supported beam experiment comparing to a modal-model-based virtual sensor, which uses modal parameters, such as mode shapes, for estimating the responses of the faulty sensors. The results show that the presented data-driven response virtual sensor technique can predict structural response with high accuracy.

  3. Convolutional Neural Network Based on Extreme Learning Machine for Maritime Ships Recognition in Infrared Images.

    Science.gov (United States)

    Khellal, Atmane; Ma, Hongbin; Fei, Qing

    2018-05-09

    The success of Deep Learning models, notably convolutional neural networks (CNNs), makes them the favorable solution for object recognition systems in both visible and infrared domains. However, the lack of training data in the case of maritime ships research leads to poor performance due to the problem of overfitting. In addition, the back-propagation algorithm used to train CNN is very slow and requires tuning many hyperparameters. To overcome these weaknesses, we introduce a new approach fully based on Extreme Learning Machine (ELM) to learn useful CNN features and perform a fast and accurate classification, which is suitable for infrared-based recognition systems. The proposed approach combines an ELM based learning algorithm to train CNN for discriminative features extraction and an ELM based ensemble for classification. The experimental results on VAIS dataset, which is the largest dataset of maritime ships, confirm that the proposed approach outperforms the state-of-the-art models in term of generalization performance and training speed. For instance, the proposed model is up to 950 times faster than the traditional back-propagation based training of convolutional neural networks, primarily for low-level features extraction.

  4. View-invariant gait recognition method by three-dimensional convolutional neural network

    Science.gov (United States)

    Xing, Weiwei; Li, Ying; Zhang, Shunli

    2018-01-01

    Gait as an important biometric feature can identify a human at a long distance. View change is one of the most challenging factors for gait recognition. To address the cross view issues in gait recognition, we propose a view-invariant gait recognition method by three-dimensional (3-D) convolutional neural network. First, 3-D convolutional neural network (3DCNN) is introduced to learn view-invariant feature, which can capture the spatial information and temporal information simultaneously on normalized silhouette sequences. Second, a network training method based on cross-domain transfer learning is proposed to solve the problem of the limited gait training samples. We choose the C3D as the basic model, which is pretrained on the Sports-1M and then fine-tune C3D model to adapt gait recognition. In the recognition stage, we use the fine-tuned model to extract gait features and use Euclidean distance to measure the similarity of gait sequences. Sufficient experiments are carried out on the CASIA-B dataset and the experimental results demonstrate that our method outperforms many other methods.

  5. Unsupervised neural spike sorting for high-density microelectrode arrays with convolutive independent component analysis.

    Science.gov (United States)

    Leibig, Christian; Wachtler, Thomas; Zeck, Günther

    2016-09-15

    Unsupervised identification of action potentials in multi-channel extracellular recordings, in particular from high-density microelectrode arrays with thousands of sensors, is an unresolved problem. While independent component analysis (ICA) achieves rapid unsupervised sorting, it ignores the convolutive structure of extracellular data, thus limiting the unmixing to a subset of neurons. Here we present a spike sorting algorithm based on convolutive ICA (cICA) to retrieve a larger number of accurately sorted neurons than with instantaneous ICA while accounting for signal overlaps. Spike sorting was applied to datasets with varying signal-to-noise ratios (SNR: 3-12) and 27% spike overlaps, sampled at either 11.5 or 23kHz on 4365 electrodes. We demonstrate how the instantaneity assumption in ICA-based algorithms has to be relaxed in order to improve the spike sorting performance for high-density microelectrode array recordings. Reformulating the convolutive mixture as an instantaneous mixture by modeling several delayed samples jointly is necessary to increase signal-to-noise ratio. Our results emphasize that different cICA algorithms are not equivalent. Spike sorting performance was assessed with ground-truth data generated from experimentally derived templates. The presented spike sorter was able to extract ≈90% of the true spike trains with an error rate below 2%. It was superior to two alternative (c)ICA methods (≈80% accurately sorted neurons) and comparable to a supervised sorting. Our new algorithm represents a fast solution to overcome the current bottleneck in spike sorting of large datasets generated by simultaneous recording with thousands of electrodes. Copyright © 2016 Elsevier B.V. All rights reserved.

  6. Transferring Deep Convolutional Neural Networks for the Scene Classification of High-Resolution Remote Sensing Imagery

    Directory of Open Access Journals (Sweden)

    Fan Hu

    2015-11-01

    Full Text Available Learning efficient image representations is at the core of the scene classification task of remote sensing imagery. The existing methods for solving the scene classification task, based on either feature coding approaches with low-level hand-engineered features or unsupervised feature learning, can only generate mid-level image features with limited representative ability, which essentially prevents them from achieving better performance. Recently, the deep convolutional neural networks (CNNs, which are hierarchical architectures trained on large-scale datasets, have shown astounding performance in object recognition and detection. However, it is still not clear how to use these deep convolutional neural networks for high-resolution remote sensing (HRRS scene classification. In this paper, we investigate how to transfer features from these successfully pre-trained CNNs for HRRS scene classification. We propose two scenarios for generating image features via extracting CNN features from different layers. In the first scenario, the activation vectors extracted from fully-connected layers are regarded as the final image features; in the second scenario, we extract dense features from the last convolutional layer at multiple scales and then encode the dense features into global image features through commonly used feature coding approaches. Extensive experiments on two public scene classification datasets demonstrate that the image features obtained by the two proposed scenarios, even with a simple linear classifier, can result in remarkable performance and improve the state-of-the-art by a significant margin. The results reveal that the features from pre-trained CNNs generalize well to HRRS datasets and are more expressive than the low- and mid-level features. Moreover, we tentatively combine features extracted from different CNN models for better performance.

  7. BrainNetCNN: Convolutional neural networks for brain networks; towards predicting neurodevelopment.

    Science.gov (United States)

    Kawahara, Jeremy; Brown, Colin J; Miller, Steven P; Booth, Brian G; Chau, Vann; Grunau, Ruth E; Zwicker, Jill G; Hamarneh, Ghassan

    2017-02-01

    We propose BrainNetCNN, a convolutional neural network (CNN) framework to predict clinical neurodevelopmental outcomes from brain networks. In contrast to the spatially local convolutions done in traditional image-based CNNs, our BrainNetCNN is composed of novel edge-to-edge, edge-to-node and node-to-graph convolutional filters that leverage the topological locality of structural brain networks. We apply the BrainNetCNN framework to predict cognitive and motor developmental outcome scores from structural brain networks of infants born preterm. Diffusion tensor images (DTI) of preterm infants, acquired between 27 and 46 weeks gestational age, were used to construct a dataset of structural brain connectivity networks. We first demonstrate the predictive capabilities of BrainNetCNN on synthetic phantom networks with simulated injury patterns and added noise. BrainNetCNN outperforms a fully connected neural-network with the same number of model parameters on both phantoms with focal and diffuse injury patterns. We then apply our method to the task of joint prediction of Bayley-III cognitive and motor scores, assessed at 18 months of age, adjusted for prematurity. We show that our BrainNetCNN framework outperforms a variety of other methods on the same data. Furthermore, BrainNetCNN is able to identify an infant's postmenstrual age to within about 2 weeks. Finally, we explore the high-level features learned by BrainNetCNN by visualizing the importance of each connection in the brain with respect to predicting the outcome scores. These findings are then discussed in the context of the anatomy and function of the developing preterm infant brain. Copyright © 2016 Elsevier Inc. All rights reserved.

  8. Fine-grained vehicle type recognition based on deep convolution neural networks

    Directory of Open Access Journals (Sweden)

    Hongcai CHEN

    2017-12-01

    Full Text Available Public security and traffic department put forward higher requirements for real-time performance and accuracy of vehicle type recognition in complex traffic scenes. Aiming at the problems of great plice forces occupation, low retrieval efficiency, and lacking of intelligence for dealing with false license, fake plate vehicles and vehicles without plates, this paper proposes a vehicle type fine-grained recognition method based GoogleNet deep convolution neural networks. The filter size and numbers of convolution neural network are designed, the activation function and vehicle type classifier are optimally selected, and a new network framework is constructed for vehicle type fine-grained recognition. The experimental results show that the proposed method has 97% accuracy for vehicle type fine-grained recognition and has greater improvement than the original GoogleNet model. Moreover, the new model effectively reduces the number of training parameters, and saves computer memory. Fine-grained vehicle type recognition can be used in intelligent traffic management area, and has important theoretical research value and practical significance.

  9. Cerebral vessels segmentation for light-sheet microscopy image using convolutional neural networks

    Science.gov (United States)

    Hu, Chaoen; Hui, Hui; Wang, Shuo; Dong, Di; Liu, Xia; Yang, Xin; Tian, Jie

    2017-03-01

    Cerebral vessel segmentation is an important step in image analysis for brain function and brain disease studies. To extract all the cerebrovascular patterns, including arteries and capillaries, some filter-based methods are used to segment vessels. However, the design of accurate and robust vessel segmentation algorithms is still challenging, due to the variety and complexity of images, especially in cerebral blood vessel segmentation. In this work, we addressed a problem of automatic and robust segmentation of cerebral micro-vessels structures in cerebrovascular images acquired by light-sheet microscope for mouse. To segment micro-vessels in large-scale image data, we proposed a convolutional neural networks (CNNs) architecture trained by 1.58 million pixels with manual label. Three convolutional layers and one fully connected layer were used in the CNNs model. We extracted a patch of size 32x32 pixels in each acquired brain vessel image as training data set to feed into CNNs for classification. This network was trained to output the probability that the center pixel of input patch belongs to vessel structures. To build the CNNs architecture, a series of mouse brain vascular images acquired from a commercial light sheet fluorescence microscopy (LSFM) system were used for training the model. The experimental results demonstrated that our approach is a promising method for effectively segmenting micro-vessels structures in cerebrovascular images with vessel-dense, nonuniform gray-level and long-scale contrast regions.

  10. Clinical Assistant Diagnosis for Electronic Medical Record Based on Convolutional Neural Network.

    Science.gov (United States)

    Yang, Zhongliang; Huang, Yongfeng; Jiang, Yiran; Sun, Yuxi; Zhang, Yu-Jin; Luo, Pengcheng

    2018-04-20

    Automatically extracting useful information from electronic medical records along with conducting disease diagnoses is a promising task for both clinical decision support(CDS) and neural language processing(NLP). Most of the existing systems are based on artificially constructed knowledge bases, and then auxiliary diagnosis is done by rule matching. In this study, we present a clinical intelligent decision approach based on Convolutional Neural Networks(CNN), which can automatically extract high-level semantic information of electronic medical records and then perform automatic diagnosis without artificial construction of rules or knowledge bases. We use collected 18,590 copies of the real-world clinical electronic medical records to train and test the proposed model. Experimental results show that the proposed model can achieve 98.67% accuracy and 96.02% recall, which strongly supports that using convolutional neural network to automatically learn high-level semantic features of electronic medical records and then conduct assist diagnosis is feasible and effective.

  11. Automatic classification of ovarian cancer types from cytological images using deep convolutional neural networks.

    Science.gov (United States)

    Wu, Miao; Yan, Chuanbo; Liu, Huiqiang; Liu, Qian

    2018-06-29

    Ovarian cancer is one of the most common gynecologic malignancies. Accurate classification of ovarian cancer types (serous carcinoma, mucous carcinoma, endometrioid carcinoma, transparent cell carcinoma) is an essential part in the different diagnosis. Computer-aided diagnosis (CADx) can provide useful advice for pathologists to determine the diagnosis correctly. In our study, we employed a Deep Convolutional Neural Networks (DCNN) based on AlexNet to automatically classify the different types of ovarian cancers from cytological images. The DCNN consists of five convolutional layers, three max pooling layers, and two full reconnect layers. Then we trained the model by two group input data separately, one was original image data and the other one was augmented image data including image enhancement and image rotation. The testing results are obtained by the method of 10-fold cross-validation, showing that the accuracy of classification models has been improved from 72.76 to 78.20% by using augmented images as training data. The developed scheme was useful for classifying ovarian cancers from cytological images. © 2018 The Author(s).

  12. LFNet: A Novel Bidirectional Recurrent Convolutional Neural Network for Light-Field Image Super-Resolution.

    Science.gov (United States)

    Wang, Yunlong; Liu, Fei; Zhang, Kunbo; Hou, Guangqi; Sun, Zhenan; Tan, Tieniu

    2018-09-01

    The low spatial resolution of light-field image poses significant difficulties in exploiting its advantage. To mitigate the dependency of accurate depth or disparity information as priors for light-field image super-resolution, we propose an implicitly multi-scale fusion scheme to accumulate contextual information from multiple scales for super-resolution reconstruction. The implicitly multi-scale fusion scheme is then incorporated into bidirectional recurrent convolutional neural network, which aims to iteratively model spatial relations between horizontally or vertically adjacent sub-aperture images of light-field data. Within the network, the recurrent convolutions are modified to be more effective and flexible in modeling the spatial correlations between neighboring views. A horizontal sub-network and a vertical sub-network of the same network structure are ensembled for final outputs via stacked generalization. Experimental results on synthetic and real-world data sets demonstrate that the proposed method outperforms other state-of-the-art methods by a large margin in peak signal-to-noise ratio and gray-scale structural similarity indexes, which also achieves superior quality for human visual systems. Furthermore, the proposed method can enhance the performance of light field applications such as depth estimation.

  13. Deep convolutional neural networks as strong gravitational lens detectors

    Science.gov (United States)

    Schaefer, C.; Geiger, M.; Kuntzer, T.; Kneib, J.-P.

    2018-03-01

    Context. Future large-scale surveys with high-resolution imaging will provide us with approximately 105 new strong galaxy-scale lenses. These strong-lensing systems will be contained in large data amounts, however, which are beyond the capacity of human experts to visually classify in an unbiased way. Aim. We present a new strong gravitational lens finder based on convolutional neural networks (CNNs). The method was applied to the strong-lensing challenge organized by the Bologna Lens Factory. It achieved first and third place, respectively, on the space-based data set and the ground-based data set. The goal was to find a fully automated lens finder for ground-based and space-based surveys that minimizes human inspection. Methods: We compared the results of our CNN architecture and three new variations ("invariant" "views" and "residual") on the simulated data of the challenge. Each method was trained separately five times on 17 000 simulated images, cross-validated using 3000 images, and then applied to a test set with 100 000 images. We used two different metrics for evaluation, the area under the receiver operating characteristic curve (AUC) score, and the recall with no false positive (Recall0FP). Results: For ground-based data, our best method achieved an AUC score of 0.977 and a Recall0FP of 0.50. For space-based data, our best method achieved an AUC score of 0.940 and a Recall0FP of 0.32. Adding dihedral invariance to the CNN architecture diminished the overall score on space-based data, but achieved a higher no-contamination recall. We found that using committees of five CNNs produced the best recall at zero contamination and consistently scored better AUC than a single CNN. Conclusions: We found that for every variation of our CNN lensfinder, we achieved AUC scores close to 1 within 6%. A deeper network did not outperform simpler CNN models either. This indicates that more complex networks are not needed to model the simulated lenses. To verify this, more

  14. Combining morphometric features and convolutional networks fusion for glaucoma diagnosis

    Science.gov (United States)

    Perdomo, Oscar; Arevalo, John; González, Fabio A.

    2017-11-01

    Glaucoma is an eye condition that leads to loss of vision and blindness. Ophthalmoscopy exam evaluates the shape, color and proportion between the optic disc and physiologic cup, but the lack of agreement among experts is still the main diagnosis problem. The application of deep convolutional neural networks combined with automatic extraction of features such as: the cup-to-disc distance in the four quadrants, the perimeter, area, eccentricity, the major radio, the minor radio in optic disc and cup, in addition to all the ratios among the previous parameters may help with a better automatic grading of glaucoma. This paper presents a strategy to merge morphological features and deep convolutional neural networks as a novel methodology to support the glaucoma diagnosis in eye fundus images.

  15. 3D Medical Image Interpolation Based on Parametric Cubic Convolution

    Institute of Scientific and Technical Information of China (English)

    2007-01-01

    In the process of display, manipulation and analysis of biomedical image data, they usually need to be converted to data of isotropic discretization through the process of interpolation, while the cubic convolution interpolation is widely used due to its good tradeoff between computational cost and accuracy. In this paper, we present a whole concept for the 3D medical image interpolation based on cubic convolution, and the six methods, with the different sharp control parameter, which are formulated in details. Furthermore, we also give an objective comparison for these methods using data sets with the different slice spacing. Each slice in these data sets is estimated by each interpolation method and compared with the original slice using three measures: mean-squared difference, number of sites of disagreement, and largest difference. According to the experimental results, we present a recommendation for 3D medical images under the different situations in the end.

  16. Spatiotemporal Recurrent Convolutional Networks for Traffic Prediction in Transportation Networks.

    Science.gov (United States)

    Yu, Haiyang; Wu, Zhihai; Wang, Shuqin; Wang, Yunpeng; Ma, Xiaolei

    2017-06-26

    Predicting large-scale transportation network traffic has become an important and challenging topic in recent decades. Inspired by the domain knowledge of motion prediction, in which the future motion of an object can be predicted based on previous scenes, we propose a network grid representation method that can retain the fine-scale structure of a transportation network. Network-wide traffic speeds are converted into a series of static images and input into a novel deep architecture, namely, spatiotemporal recurrent convolutional networks (SRCNs), for traffic forecasting. The proposed SRCNs inherit the advantages of deep convolutional neural networks (DCNNs) and long short-term memory (LSTM) neural networks. The spatial dependencies of network-wide traffic can be captured by DCNNs, and the temporal dynamics can be learned by LSTMs. An experiment on a Beijing transportation network with 278 links demonstrates that SRCNs outperform other deep learning-based algorithms in both short-term and long-term traffic prediction.

  17. ID card number detection algorithm based on convolutional neural network

    Science.gov (United States)

    Zhu, Jian; Ma, Hanjie; Feng, Jie; Dai, Leiyan

    2018-04-01

    In this paper, a new detection algorithm based on Convolutional Neural Network is presented in order to realize the fast and convenient ID information extraction in multiple scenarios. The algorithm uses the mobile device equipped with Android operating system to locate and extract the ID number; Use the special color distribution of the ID card, select the appropriate channel component; Use the image threshold segmentation, noise processing and morphological processing to take the binary processing for image; At the same time, the image rotation and projection method are used for horizontal correction when image was tilting; Finally, the single character is extracted by the projection method, and recognized by using Convolutional Neural Network. Through test shows that, A single ID number image from the extraction to the identification time is about 80ms, the accuracy rate is about 99%, It can be applied to the actual production and living environment.

  18. Trajectory Generation Method with Convolution Operation on Velocity Profile

    Energy Technology Data Exchange (ETDEWEB)

    Lee, Geon [Hanyang Univ., Seoul (Korea, Republic of); Kim, Doik [Korea Institute of Science and Technology, Daejeon (Korea, Republic of)

    2014-03-15

    The use of robots is no longer limited to the field of industrial robots and is now expanding into the fields of service and medical robots. In this light, a trajectory generation method that can respond instantaneously to the external environment is strongly required. Toward this end, this study proposes a method that enables a robot to change its trajectory in real-time using a convolution operation. The proposed method generates a trajectory in real time and satisfies the physical limits of the robot system such as acceleration and velocity limit. Moreover, a new way to improve the previous method, which generates inefficient trajectories in some cases owing to the characteristics of the trapezoidal shape of trajectories, is proposed by introducing a triangle shape. The validity and effectiveness of the proposed method is shown through a numerical simulation and a comparison with the previous convolution method.

  19. Airplane detection in remote sensing images using convolutional neural networks

    Science.gov (United States)

    Ouyang, Chao; Chen, Zhong; Zhang, Feng; Zhang, Yifei

    2018-03-01

    Airplane detection in remote sensing images remains a challenging problem and has also been taking a great interest to researchers. In this paper we propose an effective method to detect airplanes in remote sensing images using convolutional neural networks. Deep learning methods show greater advantages than the traditional methods with the rise of deep neural networks in target detection, and we give an explanation why this happens. To improve the performance on detection of airplane, we combine a region proposal algorithm with convolutional neural networks. And in the training phase, we divide the background into multi classes rather than one class, which can reduce false alarms. Our experimental results show that the proposed method is effective and robust in detecting airplane.

  20. Rock images classification by using deep convolution neural network

    Science.gov (United States)

    Cheng, Guojian; Guo, Wenhui

    2017-08-01

    Granularity analysis is one of the most essential issues in authenticate under microscope. To improve the efficiency and accuracy of traditional manual work, an convolutional neural network based method is proposed for granularity analysis from thin section image, which chooses and extracts features from image samples while build classifier to recognize granularity of input image samples. 4800 samples from Ordos basin are used for experiments under colour spaces of HSV, YCbCr and RGB respectively. On the test dataset, the correct rate in RGB colour space is 98.5%, and it is believable in HSV and YCbCr colour space. The results show that the convolution neural network can classify the rock images with high reliability.

  1. User-generated content curation with deep convolutional neural networks

    OpenAIRE

    Tous Liesa, Rubén; Wust, Otto; Gómez, Mauro; Poveda, Jonatan; Elena, Marc; Torres Viñals, Jordi; Makni, Mouna; Ayguadé Parra, Eduard

    2016-01-01

    In this paper, we report a work consisting in using deep convolutional neural networks (CNNs) for curating and filtering photos posted by social media users (Instagram and Twitter). The final goal is to facilitate searching and discovering user-generated content (UGC) with potential value for digital marketing tasks. The images are captured in real time and automatically annotated with multiple CNNs. Some of the CNNs perform generic object recognition tasks while others perform what we call v...

  2. A quantum algorithm for Viterbi decoding of classical convolutional codes

    OpenAIRE

    Grice, Jon R.; Meyer, David A.

    2014-01-01

    We present a quantum Viterbi algorithm (QVA) with better than classical performance under certain conditions. In this paper the proposed algorithm is applied to decoding classical convolutional codes, for instance; large constraint length $Q$ and short decode frames $N$. Other applications of the classical Viterbi algorithm where $Q$ is large (e.g. speech processing) could experience significant speedup with the QVA. The QVA exploits the fact that the decoding trellis is similar to the butter...

  3. Abnormality Detection in Mammography using Deep Convolutional Neural Networks

    OpenAIRE

    Xi, Pengcheng; Shu, Chang; Goubran, Rafik

    2018-01-01

    Breast cancer is the most common cancer in women worldwide. The most common screening technology is mammography. To reduce the cost and workload of radiologists, we propose a computer aided detection approach for classifying and localizing calcifications and masses in mammogram images. To improve on conventional approaches, we apply deep convolutional neural networks (CNN) for automatic feature learning and classifier building. In computer-aided mammography, deep CNN classifiers cannot be tra...

  4. Quantifying Translation-Invariance in Convolutional Neural Networks

    OpenAIRE

    Kauderer-Abrams, Eric

    2017-01-01

    A fundamental problem in object recognition is the development of image representations that are invariant to common transformations such as translation, rotation, and small deformations. There are multiple hypotheses regarding the source of translation invariance in CNNs. One idea is that translation invariance is due to the increasing receptive field size of neurons in successive convolution layers. Another possibility is that invariance is due to the pooling operation. We develop a simple ...

  5. Fast convolutional sparse coding using matrix inversion lemma

    Czech Academy of Sciences Publication Activity Database

    Šorel, Michal; Šroubek, Filip

    2016-01-01

    Roč. 55, č. 1 (2016), s. 44-51 ISSN 1051-2004 R&D Projects: GA ČR GA13-29225S Institutional support: RVO:67985556 Keywords : Convolutional sparse coding * Feature learning * Deconvolution networks * Shift-invariant sparse coding Subject RIV: JD - Computer Applications, Robotics Impact factor: 2.337, year: 2016 http://library.utia.cas.cz/separaty/2016/ZOI/sorel-0459332.pdf

  6. Shallow and deep convolutional networks for saliency prediction

    OpenAIRE

    Pan, Junting; Sayrol Clols, Elisa; Giró Nieto, Xavier; McGuinness, Kevin; O'Connor, Noel

    2016-01-01

    The prediction of salient areas in images has been traditionally addressed with hand-crafted features based on neuroscience principles. This paper, however, addresses the problem with a completely data-driven approach by training a convolutional neural network (convnet). The learning process is formulated as a minimization of a loss function that measures the Euclidean distance of the predicted saliency map with the provided ground truth. The recent publication of large datasets of saliency p...

  7. Production and reception of meaningful sound in Foville's 'encompassing convolution'.

    Science.gov (United States)

    Schiller, F

    1999-04-01

    In the history of neurology. Achille Louis Foville (1799-1879) is a name deserving to be remembered. In the course of time, his circonvolution d'enceinte of 1844 (surrounding the Sylvian fissure) became the 'convolution encompassing' every aspect of aphasiology, including amusia, ie., the localization in a coherent semicircle of semicircle of cerebral cortext serving the production and perception of language, song and instrumental music in health and disease.

  8. Relay Backpropagation for Effective Learning of Deep Convolutional Neural Networks

    OpenAIRE

    Shen, Li; Lin, Zhouchen; Huang, Qingming

    2015-01-01

    Learning deeper convolutional neural networks becomes a tendency in recent years. However, many empirical evidences suggest that performance improvement cannot be gained by simply stacking more layers. In this paper, we consider the issue from an information theoretical perspective, and propose a novel method Relay Backpropagation, that encourages the propagation of effective information through the network in training stage. By virtue of the method, we achieved the first place in ILSVRC 2015...

  9. General Dirichlet Series, Arithmetic Convolution Equations and Laplace Transforms

    Czech Academy of Sciences Publication Activity Database

    Glöckner, H.; Lucht, L.G.; Porubský, Štefan

    2009-01-01

    Roč. 193, č. 2 (2009), s. 109-129 ISSN 0039-3223 R&D Projects: GA ČR GA201/07/0191 Institutional research plan: CEZ:AV0Z10300504 Keywords : arithmetic function * Dirichlet convolution * polynomial equation * analytic equation * topological algebra * holomorphic functional calculus * implicit function theorem * Laplace transform * semigroup * complex measure Subject RIV: BA - General Mathematics Impact factor: 0.645, year: 2009 http://arxiv.org/abs/0712.3172

  10. Solving singular convolution equations using the inverse fast Fourier transform

    Czech Academy of Sciences Publication Activity Database

    Krajník, E.; Montesinos, V.; Zizler, P.; Zizler, Václav

    2012-01-01

    Roč. 57, č. 5 (2012), s. 543-550 ISSN 0862-7940 R&D Projects: GA AV ČR IAA100190901 Institutional research plan: CEZ:AV0Z10190503 Keywords : singular convolution equations * fast Fourier transform * tempered distribution Subject RIV: BA - General Mathematics Impact factor: 0.222, year: 2012 http://www.springerlink.com/content/m8437t3563214048/

  11. PSNet: prostate segmentation on MRI based on a convolutional neural network.

    Science.gov (United States)

    Tian, Zhiqiang; Liu, Lizhi; Zhang, Zhenfeng; Fei, Baowei

    2018-04-01

    Automatic segmentation of the prostate on magnetic resonance images (MRI) has many applications in prostate cancer diagnosis and therapy. We proposed a deep fully convolutional neural network (CNN) to segment the prostate automatically. Our deep CNN model is trained end-to-end in a single learning stage, which uses prostate MRI and the corresponding ground truths as inputs. The learned CNN model can be used to make an inference for pixel-wise segmentation. Experiments were performed on three data sets, which contain prostate MRI of 140 patients. The proposed CNN model of prostate segmentation (PSNet) obtained a mean Dice similarity coefficient of [Formula: see text] as compared to the manually labeled ground truth. Experimental results show that the proposed model could yield satisfactory segmentation of the prostate on MRI.

  12. Korean letter handwritten recognition using deep convolutional neural network on android platform

    Science.gov (United States)

    Purnamawati, S.; Rachmawati, D.; Lumanauw, G.; Rahmat, R. F.; Taqyuddin, R.

    2018-03-01

    Currently, popularity of Korean culture attracts many people to learn everything about Korea, particularly its language. To acquire Korean Language, every single learner needs to be able to understand Korean non-Latin character. A digital approach needs to be carried out in order to make Korean learning process easier. This study is done by using Deep Convolutional Neural Network (DCNN). DCNN performs the recognition process on the image based on the model that has been trained such as Inception-v3 Model. Subsequently, re-training process using transfer learning technique with the trained and re-trained value of model is carried though in order to develop a new model with a better performance without any specific systemic errors. The testing accuracy of this research results in 86,9%.

  13. Traffic sign recognition based on deep convolutional neural network

    Science.gov (United States)

    Yin, Shi-hao; Deng, Ji-cai; Zhang, Da-wei; Du, Jing-yuan

    2017-11-01

    Traffic sign recognition (TSR) is an important component of automated driving systems. It is a rather challenging task to design a high-performance classifier for the TSR system. In this paper, we propose a new method for TSR system based on deep convolutional neural network. In order to enhance the expression of the network, a novel structure (dubbed block-layer below) which combines network-in-network and residual connection is designed. Our network has 10 layers with parameters (block-layer seen as a single layer): the first seven are alternate convolutional layers and block-layers, and the remaining three are fully-connected layers. We train our TSR network on the German traffic sign recognition benchmark (GTSRB) dataset. To reduce overfitting, we perform data augmentation on the training images and employ a regularization method named "dropout". The activation function we employ in our network adopts scaled exponential linear units (SELUs), which can induce self-normalizing properties. To speed up the training, we use an efficient GPU to accelerate the convolutional operation. On the test dataset of GTSRB, we achieve the accuracy rate of 99.67%, exceeding the state-of-the-art results.

  14. Face recognition via Gabor and convolutional neural network

    Science.gov (United States)

    Lu, Tongwei; Wu, Menglu; Lu, Tao

    2018-04-01

    In recent years, the powerful feature learning and classification ability of convolutional neural network have attracted widely attention. Compared with the deep learning, the traditional machine learning algorithm has a good explanatory which deep learning does not have. Thus, In this paper, we propose a method to extract the feature of the traditional algorithm as the input of convolution neural network. In order to reduce the complexity of the network, the kernel function of Gabor wavelet is used to extract the feature from different position, frequency and direction of target image. It is sensitive to edge of image which can provide good direction and scale selection. The extraction of the image from eight directions on a scale are as the input of network that we proposed. The network have the advantage of weight sharing and local connection and texture feature of the input image can reduce the influence of facial expression, gesture and illumination. At the same time, we introduced a layer which combined the results of the pooling and convolution can extract deeper features. The training network used the open source caffe framework which is beneficial to feature extraction. The experiment results of the proposed method proved that the network structure effectively overcame the barrier of illumination and had a good robustness as well as more accurate and rapid than the traditional algorithm.

  15. Convolutional neural networks for the detection of diseased hearts using CT images and left atrium patches

    Science.gov (United States)

    Dormer, James D.; Halicek, Martin; Ma, Ling; Reilly, Carolyn M.; Schreibmann, Eduard; Fei, Baowei

    2018-02-01

    Cardiovascular disease is a leading cause of death in the United States. The identification of cardiac diseases on conventional three-dimensional (3D) CT can have many clinical applications. An automated method that can distinguish between healthy and diseased hearts could improve diagnostic speed and accuracy when the only modality available is conventional 3D CT. In this work, we proposed and implemented convolutional neural networks (CNNs) to identify diseased hears on CT images. Six patients with healthy hearts and six with previous cardiovascular disease events received chest CT. After the left atrium for each heart was segmented, 2D and 3D patches were created. A subset of the patches were then used to train separate convolutional neural networks using leave-one-out cross-validation of patient pairs. The results of the two neural networks were compared, with 3D patches producing the higher testing accuracy. The full list of 3D patches from the left atrium was then classified using the optimal 3D CNN model, and the receiver operating curves (ROCs) were produced. The final average area under the curve (AUC) from the ROC curves was 0.840 +/- 0.065 and the average accuracy was 78.9% +/- 5.9%. This demonstrates that the CNN-based method is capable of distinguishing healthy hearts from those with previous cardiovascular disease.

  16. Yarn-dyed fabric defect classification based on convolutional neural network

    Science.gov (United States)

    Jing, Junfeng; Dong, Amei; Li, Pengfei; Zhang, Kaibing

    2017-09-01

    Considering that manual inspection of the yarn-dyed fabric can be time consuming and inefficient, we propose a yarn-dyed fabric defect classification method by using a convolutional neural network (CNN) based on a modified AlexNet. CNN shows powerful ability in performing feature extraction and fusion by simulating the learning mechanism of human brain. The local response normalization layers in AlexNet are replaced by the batch normalization layers, which can enhance both the computational efficiency and classification accuracy. In the training process of the network, the characteristics of the defect are extracted step by step and the essential features of the image can be obtained from the fusion of the edge details with several convolution operations. Then the max-pooling layers, the dropout layers, and the fully connected layers are employed in the classification model to reduce the computation cost and extract more precise features of the defective fabric. Finally, the results of the defect classification are predicted by the softmax function. The experimental results show promising performance with an acceptable average classification rate and strong robustness on yarn-dyed fabric defect classification.

  17. Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition

    Science.gov (United States)

    Ordóñez, Francisco Javier; Roggen, Daniel

    2016-01-01

    Human activity recognition (HAR) tasks have traditionally been solved using engineered features obtained by heuristic processes. Current research suggests that deep convolutional neural networks are suited to automate feature extraction from raw sensor inputs. However, human activities are made of complex sequences of motor movements, and capturing this temporal dynamics is fundamental for successful HAR. Based on the recent success of recurrent neural networks for time series domains, we propose a generic deep framework for activity recognition based on convolutional and LSTM recurrent units, which: (i) is suitable for multimodal wearable sensors; (ii) can perform sensor fusion naturally; (iii) does not require expert knowledge in designing features; and (iv) explicitly models the temporal dynamics of feature activations. We evaluate our framework on two datasets, one of which has been used in a public activity recognition challenge. Our results show that our framework outperforms competing deep non-recurrent networks on the challenge dataset by 4% on average; outperforming some of the previous reported results by up to 9%. Our results show that the framework can be applied to homogeneous sensor modalities, but can also fuse multimodal sensors to improve performance. We characterise key architectural hyperparameters’ influence on performance to provide insights about their optimisation. PMID:26797612

  18. Residual Shuffling Convolutional Neural Networks for Deep Semantic Image Segmentation Using Multi-Modal Data

    Science.gov (United States)

    Chen, K.; Weinmann, M.; Gao, X.; Yan, M.; Hinz, S.; Jutzi, B.; Weinmann, M.

    2018-05-01

    In this paper, we address the deep semantic segmentation of aerial imagery based on multi-modal data. Given multi-modal data composed of true orthophotos and the corresponding Digital Surface Models (DSMs), we extract a variety of hand-crafted radiometric and geometric features which are provided separately and in different combinations as input to a modern deep learning framework. The latter is represented by a Residual Shuffling Convolutional Neural Network (RSCNN) combining the characteristics of a Residual Network with the advantages of atrous convolution and a shuffling operator to achieve a dense semantic labeling. Via performance evaluation on a benchmark dataset, we analyze the value of different feature sets for the semantic segmentation task. The derived results reveal that the use of radiometric features yields better classification results than the use of geometric features for the considered dataset. Furthermore, the consideration of data on both modalities leads to an improvement of the classification results. However, the derived results also indicate that the use of all defined features is less favorable than the use of selected features. Consequently, data representations derived via feature extraction and feature selection techniques still provide a gain if used as the basis for deep semantic segmentation.

  19. Detection of bars in galaxies using a deep convolutional neural network

    Science.gov (United States)

    Abraham, Sheelu; Aniyan, A. K.; Kembhavi, Ajit K.; Philip, N. S.; Vaghmare, Kaustubh

    2018-06-01

    We present an automated method for the detection of bar structure in optical images of galaxies using a deep convolutional neural network that is easy to use and provides good accuracy. In our study, we use a sample of 9346 galaxies in the redshift range of 0.009-0.2 from the Sloan Digital Sky Survey (SDSS), which has 3864 barred galaxies, the rest being unbarred. We reach a top precision of 94 per cent in identifying bars in galaxies using the trained network. This accuracy matches the accuracy reached by human experts on the same data without additional information about the images. Since deep convolutional neural networks can be scaled to handle large volumes of data, the method is expected to have great relevance in an era where astronomy data is rapidly increasing in terms of volume, variety, volatility, and velocity along with other V's that characterize big data. With the trained model, we have constructed a catalogue of barred galaxies from SDSS and made it available online.

  20. Emotional textile image classification based on cross-domain convolutional sparse autoencoders with feature selection

    Science.gov (United States)

    Li, Zuhe; Fan, Yangyu; Liu, Weihua; Yu, Zeqi; Wang, Fengqin

    2017-01-01

    We aim to apply sparse autoencoder-based unsupervised feature learning to emotional semantic analysis for textile images. To tackle the problem of limited training data, we present a cross-domain feature learning scheme for emotional textile image classification using convolutional autoencoders. We further propose a correlation-analysis-based feature selection method for the weights learned by sparse autoencoders to reduce the number of features extracted from large size images. First, we randomly collect image patches on an unlabeled image dataset in the source domain and learn local features with a sparse autoencoder. We then conduct feature selection according to the correlation between different weight vectors corresponding to the autoencoder's hidden units. We finally adopt a convolutional neural network including a pooling layer to obtain global feature activations of textile images in the target domain and send these global feature vectors into logistic regression models for emotional image classification. The cross-domain unsupervised feature learning method achieves 65% to 78% average accuracy in the cross-validation experiments corresponding to eight emotional categories and performs better than conventional methods. Feature selection can reduce the computational cost of global feature extraction by about 50% while improving classification performance.

  1. Retrieval of Sentence Sequences for an Image Stream via Coherence Recurrent Convolutional Networks.

    Science.gov (United States)

    Park, Cesc Chunseong; Kim, Youngjin; Kim, Gunhee

    2018-04-01

    We propose an approach for retrieving a sequence of natural sentences for an image stream. Since general users often take a series of pictures on their experiences, much online visual information exists in the form of image streams, for which it would better take into consideration of the whole image stream to produce natural language descriptions. While almost all previous studies have dealt with the relation between a single image and a single natural sentence, our work extends both input and output dimension to a sequence of images and a sequence of sentences. For retrieving a coherent flow of multiple sentences for a photo stream, we propose a multimodal neural architecture called coherence recurrent convolutional network (CRCN), which consists of convolutional neural networks, bidirectional long short-term memory (LSTM) networks, and an entity-based local coherence model. Our approach directly learns from vast user-generated resource of blog posts as text-image parallel training data. We collect more than 22 K unique blog posts with 170 K associated images for the travel topics of NYC, Disneyland , Australia, and Hawaii. We demonstrate that our approach outperforms other state-of-the-art image captioning methods for text sequence generation, using both quantitative measures and user studies via Amazon Mechanical Turk.

  2. Multi-scale Fully Convolutional Network for Face Detection in the Wild

    KAUST Repository

    Bai, Yancheng

    2017-08-24

    Face detection is a classical problem in computer vision. It is still a difficult task due to many nuisances that naturally occur in the wild. In this paper, we propose a multi-scale fully convolutional network for face detection. To reduce computation, the intermediate convolutional feature maps (conv) are shared by every scale model. We up-sample and down-sample the final conv map to approximate K levels of a feature pyramid, leading to a wide range of face scales that can be detected. At each feature pyramid level, a FCN is trained end-to-end to deal with faces in a small range of scale change. Because of the up-sampling, our method can detect very small faces (10×10 pixels). We test our MS-FCN detector on four public face detection datasets, including FDDB, WIDER FACE, AFW and PASCAL FACE. Extensive experiments show that it outperforms state-of-the-art methods. Also, MS-FCN runs at 23 FPS on a GPU for images of size 640×480 with no assumption on the minimum detectable face size.

  3. An Efficient Implementation of Deep Convolutional Neural Networks for MRI Segmentation.

    Science.gov (United States)

    Hoseini, Farnaz; Shahbahrami, Asadollah; Bayat, Peyman

    2018-02-27

    Image segmentation is one of the most common steps in digital image processing, classifying a digital image into different segments. The main goal of this paper is to segment brain tumors in magnetic resonance images (MRI) using deep learning. Tumors having different shapes, sizes, brightness and textures can appear anywhere in the brain. These complexities are the reasons to choose a high-capacity Deep Convolutional Neural Network (DCNN) containing more than one layer. The proposed DCNN contains two parts: architecture and learning algorithms. The architecture and the learning algorithms are used to design a network model and to optimize parameters for the network training phase, respectively. The architecture contains five convolutional layers, all using 3 × 3 kernels, and one fully connected layer. Due to the advantage of using small kernels with fold, it allows making the effect of larger kernels with smaller number of parameters and fewer computations. Using the Dice Similarity Coefficient metric, we report accuracy results on the BRATS 2016, brain tumor segmentation challenge dataset, for the complete, core, and enhancing regions as 0.90, 0.85, and 0.84 respectively. The learning algorithm includes the task-level parallelism. All the pixels of an MR image are classified using a patch-based approach for segmentation. We attain a good performance and the experimental results show that the proposed DCNN increases the segmentation accuracy compared to previous techniques.

  4. AUC-Maximized Deep Convolutional Neural Fields for Protein Sequence Labeling.

    Science.gov (United States)

    Wang, Sheng; Sun, Siqi; Xu, Jinbo

    2016-09-01

    Deep Convolutional Neural Networks (DCNN) has shown excellent performance in a variety of machine learning tasks. This paper presents Deep Convolutional Neural Fields (DeepCNF), an integration of DCNN with Conditional Random Field (CRF), for sequence labeling with an imbalanced label distribution. The widely-used training methods, such as maximum-likelihood and maximum labelwise accuracy, do not work well on imbalanced data. To handle this, we present a new training algorithm called maximum-AUC for DeepCNF. That is, we train DeepCNF by directly maximizing the empirical Area Under the ROC Curve (AUC), which is an unbiased measurement for imbalanced data. To fulfill this, we formulate AUC in a pairwise ranking framework, approximate it by a polynomial function and then apply a gradient-based procedure to optimize it. Our experimental results confirm that maximum-AUC greatly outperforms the other two training methods on 8-state secondary structure prediction and disorder prediction since their label distributions are highly imbalanced and also has similar performance as the other two training methods on solvent accessibility prediction, which has three equally-distributed labels. Furthermore, our experimental results show that our AUC-trained DeepCNF models greatly outperform existing popular predictors of these three tasks. The data and software related to this paper are available at https://github.com/realbigws/DeepCNF_AUC.

  5. Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition

    Directory of Open Access Journals (Sweden)

    Francisco Javier Ordóñez

    2016-01-01

    Full Text Available Human activity recognition (HAR tasks have traditionally been solved using engineered features obtained by heuristic processes. Current research suggests that deep convolutional neural networks are suited to automate feature extraction from raw sensor inputs. However, human activities are made of complex sequences of motor movements, and capturing this temporal dynamics is fundamental for successful HAR. Based on the recent success of recurrent neural networks for time series domains, we propose a generic deep framework for activity recognition based on convolutional and LSTM recurrent units, which: (i is suitable for multimodal wearable sensors; (ii can perform sensor fusion naturally; (iii does not require expert knowledge in designing features; and (iv explicitly models the temporal dynamics of feature activations. We evaluate our framework on two datasets, one of which has been used in a public activity recognition challenge. Our results show that our framework outperforms competing deep non-recurrent networks on the challenge dataset by 4% on average; outperforming some of the previous reported results by up to 9%. Our results show that the framework can be applied to homogeneous sensor modalities, but can also fuse multimodal sensors to improve performance. We characterise key architectural hyperparameters’ influence on performance to provide insights about their optimisation.

  6. Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition.

    Science.gov (United States)

    Ordóñez, Francisco Javier; Roggen, Daniel

    2016-01-18

    Human activity recognition (HAR) tasks have traditionally been solved using engineered features obtained by heuristic processes. Current research suggests that deep convolutional neural networks are suited to automate feature extraction from raw sensor inputs. However, human activities are made of complex sequences of motor movements, and capturing this temporal dynamics is fundamental for successful HAR. Based on the recent success of recurrent neural networks for time series domains, we propose a generic deep framework for activity recognition based on convolutional and LSTM recurrent units, which: (i) is suitable for multimodal wearable sensors; (ii) can perform sensor fusion naturally; (iii) does not require expert knowledge in designing features; and (iv) explicitly models the temporal dynamics of feature activations. We evaluate our framework on two datasets, one of which has been used in a public activity recognition challenge. Our results show that our framework outperforms competing deep non-recurrent networks on the challenge dataset by 4% on average; outperforming some of the previous reported results by up to 9%. Our results show that the framework can be applied to homogeneous sensor modalities, but can also fuse multimodal sensors to improve performance. We characterise key architectural hyperparameters' influence on performance to provide insights about their optimisation.

  7. Infrared dim moving target tracking via sparsity-based discriminative classifier and convolutional network

    Science.gov (United States)

    Qian, Kun; Zhou, Huixin; Wang, Bingjian; Song, Shangzhen; Zhao, Dong

    2017-11-01

    Infrared dim and small target tracking is a great challenging task. The main challenge for target tracking is to account for appearance change of an object, which submerges in the cluttered background. An efficient appearance model that exploits both the global template and local representation over infrared image sequences is constructed for dim moving target tracking. A Sparsity-based Discriminative Classifier (SDC) and a Convolutional Network-based Generative Model (CNGM) are combined with a prior model. In the SDC model, a sparse representation-based algorithm is adopted to calculate the confidence value that assigns more weights to target templates than negative background templates. In the CNGM model, simple cell feature maps are obtained by calculating the convolution between target templates and fixed filters, which are extracted from the target region at the first frame. These maps measure similarities between each filter and local intensity patterns across the target template, therefore encoding its local structural information. Then, all the maps form a representation, preserving the inner geometric layout of a candidate template. Furthermore, the fixed target template set is processed via an efficient prior model. The same operation is applied to candidate templates in the CNGM model. The online update scheme not only accounts for appearance variations but also alleviates the migration problem. At last, collaborative confidence values of particles are utilized to generate particles' importance weights. Experiments on various infrared sequences have validated the tracking capability of the presented algorithm. Experimental results show that this algorithm runs in real-time and provides a higher accuracy than state of the art algorithms.

  8. Relative location prediction in CT scan images using convolutional neural networks.

    Science.gov (United States)

    Guo, Jiajia; Du, Hongwei; Zhu, Jianyue; Yan, Ting; Qiu, Bensheng

    2018-07-01

    Relative location prediction in computed tomography (CT) scan images is a challenging problem. Many traditional machine learning methods have been applied in attempts to alleviate this problem. However, the accuracy and speed of these methods cannot meet the requirement of medical scenario. In this paper, we propose a regression model based on one-dimensional convolutional neural networks (CNN) to determine the relative location of a CT scan image both quickly and precisely. In contrast to other common CNN models that use a two-dimensional image as an input, the input of this CNN model is a feature vector extracted by a shape context algorithm with spatial correlation. Normalization via z-score is first applied as a pre-processing step. Then, in order to prevent overfitting and improve model's performance, 20% of the elements of the feature vectors are randomly set to zero. This CNN model consists primarily of three one-dimensional convolutional layers, three dropout layers and two fully-connected layers with appropriate loss functions. A public dataset is employed to validate the performance of the proposed model using a 5-fold cross validation. Experimental results demonstrate an excellent performance of the proposed model when compared with contemporary techniques, achieving a median absolute error of 1.04 cm and mean absolute error of 1.69 cm. The time taken for each relative location prediction is approximately 2 ms. Results indicate that the proposed CNN method can contribute to a quick and accurate relative location prediction in CT scan images, which can improve efficiency of the medical picture archiving and communication system in the future. Copyright © 2018 Elsevier B.V. All rights reserved.

  9. Bioprinting of 3D Convoluted Renal Proximal Tubules on Perfusable Chips

    Science.gov (United States)

    Homan, Kimberly A.; Kolesky, David B.; Skylar-Scott, Mark A.; Herrmann, Jessica; Obuobi, Humphrey; Moisan, Annie; Lewis, Jennifer A.

    2016-10-01

    Three-dimensional models of kidney tissue that recapitulate human responses are needed for drug screening, disease modeling, and, ultimately, kidney organ engineering. Here, we report a bioprinting method for creating 3D human renal proximal tubules in vitro that are fully embedded within an extracellular matrix and housed in perfusable tissue chips, allowing them to be maintained for greater than two months. Their convoluted tubular architecture is circumscribed by proximal tubule epithelial cells and actively perfused through the open lumen. These engineered 3D proximal tubules on chip exhibit significantly enhanced epithelial morphology and functional properties relative to the same cells grown on 2D controls with or without perfusion. Upon introducing the nephrotoxin, Cyclosporine A, the epithelial barrier is disrupted in a dose-dependent manner. Our bioprinting method provides a new route for programmably fabricating advanced human kidney tissue models on demand.

  10. Super-resolution using a light inception layer in convolutional neural network

    Science.gov (United States)

    Mou, Qinyang; Guo, Jun

    2018-04-01

    Recently, several models based on CNN architecture have achieved great result on Single Image Super-Resolution (SISR) problem. In this paper, we propose an image super-resolution method (SR) using a light inception layer in convolutional network (LICN). Due to the strong representation ability of our well-designed inception layer that can learn richer representation with less parameters, we can build our model with shallow architecture that can reduce the effect of vanishing gradients problem and save computational costs. Our model strike a balance between computational speed and the quality of the result. Compared with state-of-the-art result, we produce comparable or better results with faster computational speed.

  11. Segmentation of left ventricle myocardium in porcine cardiac cine MR images using a hybrid of fully convolutional neural networks and convolutional LSTM

    Science.gov (United States)

    Zhang, Dongqing; Icke, Ilknur; Dogdas, Belma; Parimal, Sarayu; Sampath, Smita; Forbes, Joseph; Bagchi, Ansuman; Chin, Chih-Liang; Chen, Antong

    2018-03-01

    In the development of treatments for cardiovascular diseases, short axis cardiac cine MRI is important for the assessment of various structural and functional properties of the heart. In short axis cardiac cine MRI, Cardiac properties including the ventricle dimensions, stroke volume, and ejection fraction can be extracted based on accurate segmentation of the left ventricle (LV) myocardium. One of the most advanced segmentation methods is based on fully convolutional neural networks (FCN) and can be successfully used to do segmentation in cardiac cine MRI slices. However, the temporal dependency between slices acquired at neighboring time points is not used. Here, based on our previously proposed FCN structure, we proposed a new algorithm to segment LV myocardium in porcine short axis cardiac cine MRI by incorporating convolutional long short-term memory (Conv-LSTM) to leverage the temporal dependency. In this approach, instead of processing each slice independently in a conventional CNN-based approach, the Conv-LSTM architecture captures the dynamics of cardiac motion over time. In a leave-one-out experiment on 8 porcine specimens (3,600 slices), the proposed approach was shown to be promising by achieving average mean Dice similarity coefficient (DSC) of 0.84, Hausdorff distance (HD) of 6.35 mm, and average perpendicular distance (APD) of 1.09 mm when compared with manual segmentations, which improved the performance of our previous FCN-based approach (average mean DSC=0.84, HD=6.78 mm, and APD=1.11 mm). Qualitatively, our model showed robustness against low image quality and complications in the surrounding anatomy due to its ability to capture the dynamics of cardiac motion.

  12. Convolution equations on lattices: periodic solutions with values in a prime characteristic field

    OpenAIRE

    Zaidenberg, Mikhail

    2006-01-01

    These notes are inspired by the theory of cellular automata. A linear cellular automaton on a lattice of finite rank or on a toric grid is a discrete dinamical system generated by a convolution operator with kernel concentrated in the nearest neighborhood of the origin. In the present paper we deal with general convolution operators. We propose an approach via harmonic analysis which works over a field of positive characteristic. It occurs that a standard spectral problem for a convolution op...

  13. Automatic sleep stage classification of single-channel EEG by using complex-valued convolutional neural network.

    Science.gov (United States)

    Zhang, Junming; Wu, Yan

    2018-03-28

    Many systems are developed for automatic sleep stage classification. However, nearly all models are based on handcrafted features. Because of the large feature space, there are so many features that feature selection should be used. Meanwhile, designing handcrafted features is a difficult and time-consuming task because the feature designing needs domain knowledge of experienced experts. Results vary when different sets of features are chosen to identify sleep stages. Additionally, many features that we may be unaware of exist. However, these features may be important for sleep stage classification. Therefore, a new sleep stage classification system, which is based on the complex-valued convolutional neural network (CCNN), is proposed in this study. Unlike the existing sleep stage methods, our method can automatically extract features from raw electroencephalography data and then classify sleep stage based on the learned features. Additionally, we also prove that the decision boundaries for the real and imaginary parts of a complex-valued convolutional neuron intersect orthogonally. The classification performances of handcrafted features are compared with those of learned features via CCNN. Experimental results show that the proposed method is comparable to the existing methods. CCNN obtains a better classification performance and considerably faster convergence speed than convolutional neural network. Experimental results also show that the proposed method is a useful decision-support tool for automatic sleep stage classification.

  14. Performance Analysis of DPSK Signals with Selection Combining and Convolutional Coding in Fading Channel

    National Research Council Canada - National Science Library

    Ong, Choon

    1998-01-01

    The performance analysis of a differential phase shift keyed (DPSK) communications system, operating in a Rayleigh fading environment, employing convolutional coding and diversity processing is presented...

  15. Classification of CITES-listed and other neotropical Meliaceae wood images using convolutional neural networks.

    Science.gov (United States)

    Ravindran, Prabu; Costa, Adriana; Soares, Richard; Wiedenhoeft, Alex C

    2018-01-01

    The current state-of-the-art for field wood identification to combat illegal logging relies on experienced practitioners using hand lenses, specialized identification keys, atlases of woods, and field manuals. Accumulation of this expertise is time-consuming and access to training is relatively rare compared to the international demand for field wood identification. A reliable, consistent and cost effective field screening method is necessary for effective global scale enforcement of international treaties such as the Convention on the International Trade in Endagered Species (CITES) or national laws (e.g. the US Lacey Act) governing timber trade and imports. We present highly effective computer vision classification models, based on deep convolutional neural networks, trained via transfer learning, to identify the woods of 10 neotropical species in the family Meliaceae, including CITES-listed Swietenia macrophylla , Swietenia mahagoni , Cedrela fissilis , and Cedrela odorata . We build and evaluate models to classify the 10 woods at the species and genus levels, with image-level model accuracy ranging from 87.4 to 97.5%, with the strongest performance by the genus-level model. Misclassified images are attributed to classes consistent with traditional wood anatomical results, and our species-level accuracy greatly exceeds the resolution of traditional wood identification. The end-to-end trained image classifiers that we present discriminate the woods based on digital images of the transverse surface of solid wood blocks, which are surfaces and images that can be prepared and captured in the field. Hence this work represents a strong proof-of-concept for using computer vision and convolutional neural networks to develop practical models for field screening timber and wood products to combat illegal logging.

  16. An adaptive deep convolutional neural network for rolling bearing fault diagnosis

    International Nuclear Information System (INIS)

    Fuan, Wang; Hongkai, Jiang; Haidong, Shao; Wenjing, Duan; Shuaipeng, Wu

    2017-01-01

    The working conditions of rolling bearings usually is very complex, which makes it difficult to diagnose rolling bearing faults. In this paper, a novel method called the adaptive deep convolutional neural network (CNN) is proposed for rolling bearing fault diagnosis. Firstly, to get rid of manual feature extraction, the deep CNN model is initialized for automatic feature learning. Secondly, to adapt to different signal characteristics, the main parameters of the deep CNN model are determined with a particle swarm optimization method. Thirdly, to evaluate the feature learning ability of the proposed method, t-distributed stochastic neighbor embedding (t-SNE) is further adopted to visualize the hierarchical feature learning process. The proposed method is applied to diagnose rolling bearing faults, and the results confirm that the proposed method is more effective and robust than other intelligent methods. (paper)

  17. ELHnet: a convolutional neural network for classifying cochlear endolymphatic hydrops imaged with optical coherence tomography.

    Science.gov (United States)

    Liu, George S; Zhu, Michael H; Kim, Jinkyung; Raphael, Patrick; Applegate, Brian E; Oghalai, John S

    2017-10-01

    Detection of endolymphatic hydrops is important for diagnosing Meniere's disease, and can be performed non-invasively using optical coherence tomography (OCT) in animal models as well as potentially in the clinic. Here, we developed ELHnet, a convolutional neural network to classify endolymphatic hydrops in a mouse model using learned features from OCT images of mice cochleae. We trained ELHnet on 2159 training and validation images from 17 mice, using only the image pixels and observer-determined labels of endolymphatic hydrops as the inputs. We tested ELHnet on 37 images from 37 mice that were previously not used, and found that the neural network correctly classified 34 of the 37 mice. This demonstrates an improvement in performance from previous work on computer-aided classification of endolymphatic hydrops. To the best of our knowledge, this is the first deep CNN designed for endolymphatic hydrops classification.

  18. Constrained Convolutional Sparse Coding for Parametric Based Reconstruction of Line Drawings

    KAUST Repository

    Shaheen, Sara

    2017-12-25

    Convolutional sparse coding (CSC) plays an essential role in many computer vision applications ranging from image compression to deep learning. In this work, we spot the light on a new application where CSC can effectively serve, namely line drawing analysis. The process of drawing a line drawing can be approximated as the sparse spatial localization of a number of typical basic strokes, which in turn can be cast as a non-standard CSC model that considers the line drawing formation process from parametric curves. These curves are learned to optimize the fit between the model and a specific set of line drawings. Parametric representation of sketches is vital in enabling automatic sketch analysis, synthesis and manipulation. A couple of sketch manipulation examples are demonstrated in this work. Consequently, our novel method is expected to provide a reliable and automatic method for parametric sketch description. Through experiments, we empirically validate the convergence of our method to a feasible solution.

  19. Rolling bearing fault feature learning using improved convolutional deep belief network with compressed sensing

    Science.gov (United States)

    Shao, Haidong; Jiang, Hongkai; Zhang, Haizhou; Duan, Wenjing; Liang, Tianchen; Wu, Shuaipeng

    2018-02-01

    The vibration signals collected from rolling bearing are usually complex and non-stationary with heavy background noise. Therefore, it is a great challenge to efficiently learn the representative fault features of the collected vibration signals. In this paper, a novel method called improved convolutional deep belief network (CDBN) with compressed sensing (CS) is developed for feature learning and fault diagnosis of rolling bearing. Firstly, CS is adopted for reducing the vibration data amount to improve analysis efficiency. Secondly, a new CDBN model is constructed with Gaussian visible units to enhance the feature learning ability for the compressed data. Finally, exponential moving average (EMA) technique is employed to improve the generalization performance of the constructed deep model. The developed method is applied to analyze the experimental rolling bearing vibration signals. The results confirm that the developed method is more effective than the traditional methods.

  20. Composability-Centered Convolutional Neural Network Pruning

    Energy Technology Data Exchange (ETDEWEB)

    Shen, Xipeng [North Carolina State University; Guan, Hui [North Carolina State University; Lim, Seung-Hwan [ORNL; Patton, Robert M. [ORNL

    2018-02-01

    This work studies the composability of the building blocks ofstructural CNN models (e.g., GoogleLeNet and Residual Networks) in thecontext of network pruning. We empirically validate that a networkcomposed of pre-trained building blocks (e.g. residual blocks andInception modules) not only gives a better initial setting fortraining, but also allows the training process to converge at asignificantly higher accuracy in much less time. Based on thatinsight, we propose a {\\em composability-centered} design for CNNnetwork pruning. Experiments show that this new scheme shortens theconfiguration process in CNN network pruning by up to 186.8X forResNet-50 and up to 30.2X for Inception-V3, and meanwhile, the modelsit finds that meet the accuracy requirement are significantly morecompact than those found by default schemes.

  1. Biased Dropout and Crossmap Dropout: Learning towards effective Dropout regularization in convolutional neural network.

    Science.gov (United States)

    Poernomo, Alvin; Kang, Dae-Ki

    2018-08-01

    Training a deep neural network with a large number of parameters often leads to overfitting problem. Recently, Dropout has been introduced as a simple, yet effective regularization approach to combat overfitting in such models. Although Dropout has shown remarkable results on many deep neural network cases, its actual effect on CNN has not been thoroughly explored. Moreover, training a Dropout model will significantly increase the training time as it takes longer time to converge than a non-Dropout model with the same architecture. To deal with these issues, we address Biased Dropout and Crossmap Dropout, two novel approaches of Dropout extension based on the behavior of hidden units in CNN model. Biased Dropout divides the hidden units in a certain layer into two groups based on their magnitude and applies different Dropout rate to each group appropriately. Hidden units with higher activation value, which give more contributions to the network final performance, will be retained by a lower Dropout rate, while units with lower activation value will be exposed to a higher Dropout rate to compensate the previous part. The second approach is Crossmap Dropout, which is an extension of the regular Dropout in convolution layer. Each feature map in a convolution layer has a strong correlation between each other, particularly in every identical pixel location in each feature map. Crossmap Dropout tries to maintain this important correlation yet at the same time break the correlation between each adjacent pixel with respect to all feature maps by applying the same Dropout mask to all feature maps, so that all pixels or units in equivalent positions in each feature map will be either dropped or active during training. Our experiment with various benchmark datasets shows that our approaches provide better generalization than the regular Dropout. Moreover, our Biased Dropout takes faster time to converge during training phase, suggesting that assigning noise appropriately in

  2. Finding strong lenses in CFHTLS using convolutional neural networks

    Science.gov (United States)

    Jacobs, C.; Glazebrook, K.; Collett, T.; More, A.; McCarthy, C.

    2017-10-01

    We train and apply convolutional neural networks, a machine learning technique developed to learn from and classify image data, to Canada-France-Hawaii Telescope Legacy Survey (CFHTLS) imaging for the identification of potential strong lensing systems. An ensemble of four convolutional neural networks was trained on images of simulated galaxy-galaxy lenses. The training sets consisted of a total of 62 406 simulated lenses and 64 673 non-lens negative examples generated with two different methodologies. An ensemble of trained networks was applied to all of the 171 deg2 of the CFHTLS wide field image data, identifying 18 861 candidates including 63 known and 139 other potential lens candidates. A second search of 1.4 million early-type galaxies selected from the survey catalogue as potential deflectors, identified 2465 candidates including 117 previously known lens candidates, 29 confirmed lenses/high-quality lens candidates, 266 novel probable or potential lenses and 2097 candidates we classify as false positives. For the catalogue-based search we estimate a completeness of 21-28 per cent with respect to detectable lenses and a purity of 15 per cent, with a false-positive rate of 1 in 671 images tested. We predict a human astronomer reviewing candidates produced by the system would identify 20 probable lenses and 100 possible lenses per hour in a sample selected by the robot. Convolutional neural networks are therefore a promising tool for use in the search for lenses in current and forthcoming surveys such as the Dark Energy Survey and the Large Synoptic Survey Telescope.

  3. Fast space-varying convolution using matrix source coding with applications to camera stray light reduction.

    Science.gov (United States)

    Wei, Jianing; Bouman, Charles A; Allebach, Jan P

    2014-05-01

    Many imaging applications require the implementation of space-varying convolution for accurate restoration and reconstruction of images. Here, we use the term space-varying convolution to refer to linear operators whose impulse response has slow spatial variation. In addition, these space-varying convolution operators are often dense, so direct implementation of the convolution operator is typically computationally impractical. One such example is the problem of stray light reduction in digital cameras, which requires the implementation of a dense space-varying deconvolution operator. However, other inverse problems, such as iterative tomographic reconstruction, can also depend on the implementation of dense space-varying convolution. While space-invariant convolution can be efficiently implemented with the fast Fourier transform, this approach does not work for space-varying operators. So direct convolution is often the only option for implementing space-varying convolution. In this paper, we develop a general approach to the efficient implementation of space-varying convolution, and demonstrate its use in the application of stray light reduction. Our approach, which we call matrix source coding, is based on lossy source coding of the dense space-varying convolution matrix. Importantly, by coding the transformation matrix, we not only reduce the memory required to store it; we also dramatically reduce the computation required to implement matrix-vector products. Our algorithm is able to reduce computation by approximately factoring the dense space-varying convolution operator into a product of sparse transforms. Experimental results show that our method can dramatically reduce the computation required for stray light reduction while maintaining high accuracy.

  4. Fast Convolutional Sparse Coding in the Dual Domain

    KAUST Repository

    Affara, Lama Ahmed

    2017-09-27

    Convolutional sparse coding (CSC) is an important building block of many computer vision applications ranging from image and video compression to deep learning. We present two contributions to the state of the art in CSC. First, we significantly speed up the computation by proposing a new optimization framework that tackles the problem in the dual domain. Second, we extend the original formulation to higher dimensions in order to process a wider range of inputs, such as color inputs, or HOG features. Our results show a significant speedup compared to the current state of the art in CSC.

  5. Phase transitions in glassy systems via convolutional neural networks

    Science.gov (United States)

    Fang, Chao

    Machine learning is a powerful approach commonplace in industry to tackle large data sets. Most recently, it has found its way into condensed matter physics, allowing for the first time the study of, e.g., topological phase transitions and strongly-correlated electron systems. The study of spin glasses is plagued by finite-size effects due to the long thermalization times needed. Here we use convolutional neural networks in an attempt to detect a phase transition in three-dimensional Ising spin glasses. Our results are compared to traditional approaches.

  6. Visualizing Vector Fields Using Line Integral Convolution and Dye Advection

    Science.gov (United States)

    Shen, Han-Wei; Johnson, Christopher R.; Ma, Kwan-Liu

    1996-01-01

    We present local and global techniques to visualize three-dimensional vector field data. Using the Line Integral Convolution (LIC) method to image the global vector field, our new algorithm allows the user to introduce colored 'dye' into the vector field to highlight local flow features. A fast algorithm is proposed that quickly recomputes the dyed LIC images. In addition, we introduce volume rendering methods that can map the LIC texture on any contour surface and/or translucent region defined by additional scalar quantities, and can follow the advection of colored dye throughout the volume.

  7. Fast Convolutional Sparse Coding in the Dual Domain

    KAUST Repository

    Affara, Lama Ahmed; Ghanem, Bernard; Wonka, Peter

    2017-01-01

    Convolutional sparse coding (CSC) is an important building block of many computer vision applications ranging from image and video compression to deep learning. We present two contributions to the state of the art in CSC. First, we significantly speed up the computation by proposing a new optimization framework that tackles the problem in the dual domain. Second, we extend the original formulation to higher dimensions in order to process a wider range of inputs, such as color inputs, or HOG features. Our results show a significant speedup compared to the current state of the art in CSC.

  8. Tandem mass spectrometry data quality assessment by self-convolution

    Directory of Open Access Journals (Sweden)

    Tham Wai

    2007-09-01

    Full Text Available Abstract Background Many algorithms have been developed for deciphering the tandem mass spectrometry (MS data sets. They can be essentially clustered into two classes. The first performs searches on theoretical mass spectrum database, while the second based itself on de novo sequencing from raw mass spectrometry data. It was noted that the quality of mass spectra affects significantly the protein identification processes in both instances. This prompted the authors to explore ways to measure the quality of MS data sets before subjecting them to the protein identification algorithms, thus allowing for more meaningful searches and increased confidence level of proteins identified. Results The proposed method measures the qualities of MS data sets based on the symmetric property of b- and y-ion peaks present in a MS spectrum. Self-convolution on MS data and its time-reversal copy was employed. Due to the symmetric nature of b-ions and y-ions peaks, the self-convolution result of a good spectrum would produce a highest mid point intensity peak. To reduce processing time, self-convolution was achieved using Fast Fourier Transform and its inverse transform, followed by the removal of the "DC" (Direct Current component and the normalisation of the data set. The quality score was defined as the ratio of the intensity at the mid point to the remaining peaks of the convolution result. The method was validated using both theoretical mass spectra, with various permutations, and several real MS data sets. The results were encouraging, revealing a high percentage of positive prediction rates for spectra with good quality scores. Conclusion We have demonstrated in this work a method for determining the quality of tandem MS data set. By pre-determining the quality of tandem MS data before subjecting them to protein identification algorithms, spurious protein predictions due to poor tandem MS data are avoided, giving scientists greater confidence in the

  9. The Use of Finite Fields and Rings to Compute Convolutions

    Science.gov (United States)

    1975-06-06

    showed in Ref. 1 that the convolution of two finite sequences of integers (a, ) and (b, ) for k = 1, 2, . . ., d can be obtained as the inverse transform of...since the T.’S are all distinct. Thus T~ exists and (7) can be solved as a = T A the inverse " transform .𔃻 Next let us impose on (7) the...the inverse transform d-1 Cn= (d) I Cka k=0 If an a can be found so that multiplications by powers of a are simple in hardware, the

  10. Tandem mass spectrometry data quality assessment by self-convolution.

    Science.gov (United States)

    Choo, Keng Wah; Tham, Wai Mun

    2007-09-20

    Many algorithms have been developed for deciphering the tandem mass spectrometry (MS) data sets. They can be essentially clustered into two classes. The first performs searches on theoretical mass spectrum database, while the second based itself on de novo sequencing from raw mass spectrometry data. It was noted that the quality of mass spectra affects significantly the protein identification processes in both instances. This prompted the authors to explore ways to measure the quality of MS data sets before subjecting them to the protein identification algorithms, thus allowing for more meaningful searches and increased confidence level of proteins identified. The proposed method measures the qualities of MS data sets based on the symmetric property of b- and y-ion peaks present in a MS spectrum. Self-convolution on MS data and its time-reversal copy was employed. Due to the symmetric nature of b-ions and y-ions peaks, the self-convolution result of a good spectrum would produce a highest mid point intensity peak. To reduce processing time, self-convolution was achieved using Fast Fourier Transform and its inverse transform, followed by the removal of the "DC" (Direct Current) component and the normalisation of the data set. The quality score was defined as the ratio of the intensity at the mid point to the remaining peaks of the convolution result. The method was validated using both theoretical mass spectra, with various permutations, and several real MS data sets. The results were encouraging, revealing a high percentage of positive prediction rates for spectra with good quality scores. We have demonstrated in this work a method for determining the quality of tandem MS data set. By pre-determining the quality of tandem MS data before subjecting them to protein identification algorithms, spurious protein predictions due to poor tandem MS data are avoided, giving scientists greater confidence in the predicted results. We conclude that the algorithm performs well

  11. Weed Growth Stage Estimator Using Deep Convolutional Neural Networks

    DEFF Research Database (Denmark)

    Teimouri, Nima; Dyrmann, Mads; Nielsen, Per Rydahl

    2018-01-01

    This study outlines a new method of automatically estimating weed species and growth stages (from cotyledon until eight leaves are visible) of in situ images covering 18 weed species or families. Images of weeds growing within a variety of crops were gathered across variable environmental conditi...... in estimating the number of leaves and 96% accuracy when accepting a deviation of two leaves. These results show that this new method of using deep convolutional neural networks has a relatively high ability to estimate early growth stages across a wide variety of weed species....

  12. DNCON2: improved protein contact prediction using two-level deep convolutional neural networks.

    Science.gov (United States)

    Adhikari, Badri; Hou, Jie; Cheng, Jianlin

    2018-05-01

    Significant improvements in the prediction of protein residue-residue contacts are observed in the recent years. These contacts, predicted using a variety of coevolution-based and machine learning methods, are the key contributors to the recent progress in ab initio protein structure prediction, as demonstrated in the recent CASP experiments. Continuing the development of new methods to reliably predict contact maps is essential to further improve ab initio structure prediction. In this paper we discuss DNCON2, an improved protein contact map predictor based on two-level deep convolutional neural networks. It consists of six convolutional neural networks-the first five predict contacts at 6, 7.5, 8, 8.5 and 10 Å distance thresholds, and the last one uses these five predictions as additional features to predict final contact maps. On the free-modeling datasets in CASP10, 11 and 12 experiments, DNCON2 achieves mean precisions of 35, 50 and 53.4%, respectively, higher than 30.6% by MetaPSICOV on CASP10 dataset, 34% by MetaPSICOV on CASP11 dataset and 46.3% by Raptor-X on CASP12 dataset, when top L/5 long-range contacts are evaluated. We attribute the improved performance of DNCON2 to the inclusion of short- and medium-range contacts into training, two-level approach to prediction, use of the state-of-the-art optimization and activation functions, and a novel deep learning architecture that allows each filter in a convolutional layer to access all the input features of a protein of arbitrary length. The web server of DNCON2 is at http://sysbio.rnet.missouri.edu/dncon2/ where training and testing datasets as well as the predictions for CASP10, 11 and 12 free-modeling datasets can also be downloaded. Its source code is available at https://github.com/multicom-toolbox/DNCON2/. chengji@missouri.edu. Supplementary data are available at Bioinformatics online.

  13. A MacWilliams Identity for Convolutional Codes: The General Case

    OpenAIRE

    Gluesing-Luerssen, Heide; Schneider, Gert

    2008-01-01

    A MacWilliams Identity for convolutional codes will be established. It makes use of the weight adjacency matrices of the code and its dual, based on state space realizations (the controller canonical form) of the codes in question. The MacWilliams Identity applies to various notions of duality appearing in the literature on convolutional coding theory.

  14. Isointense infant brain MRI segmentation with a dilated convolutional neural network

    NARCIS (Netherlands)

    Moeskops, P.; Pluim, J.P.W.

    2017-01-01

    Quantitative analysis of brain MRI at the age of 6 months is difficult because of the limited contrast between white matter and gray matter. In this study, we use a dilated triplanar convolutional neural network in combination with a non-dilated 3D convolutional neural network for the segmentation

  15. An upper bound on the number of errors corrected by a convolutional code

    DEFF Research Database (Denmark)

    Justesen, Jørn

    2000-01-01

    The number of errors that a convolutional codes can correct in a segment of the encoded sequence is upper bounded by the number of distinct syndrome sequences of the relevant length.......The number of errors that a convolutional codes can correct in a segment of the encoded sequence is upper bounded by the number of distinct syndrome sequences of the relevant length....

  16. Linear diffusion-wave channel routing using a discrete Hayami convolution method

    Science.gov (United States)

    Li Wang; Joan Q. Wu; William J. Elliot; Fritz R. Feidler; Sergey. Lapin

    2014-01-01

    The convolution of an input with a response function has been widely used in hydrology as a means to solve various problems analytically. Due to the high computation demand in solving the functions using numerical integration, it is often advantageous to use the discrete convolution instead of the integration of the continuous functions. This approach greatly reduces...

  17. Using convolutional decoding to improve time delay and phase estimation in digital communications

    Science.gov (United States)

    Ormesher, Richard C [Albuquerque, NM; Mason, John J [Albuquerque, NM

    2010-01-26

    The time delay and/or phase of a communication signal received by a digital communication receiver can be estimated based on a convolutional decoding operation that the communication receiver performs on the received communication signal. If the original transmitted communication signal has been spread according to a spreading operation, a corresponding despreading operation can be integrated into the convolutional decoding operation.

  18. Enhancing Hi-C data resolution with deep convolutional neural network HiCPlus.

    Science.gov (United States)

    Zhang, Yan; An, Lin; Xu, Jie; Zhang, Bo; Zheng, W Jim; Hu, Ming; Tang, Jijun; Yue, Feng

    2018-02-21

    Although Hi-C technology is one of the most popular tools for studying 3D genome organization, due to sequencing cost, the resolution of most Hi-C datasets are coarse and cannot be used to link distal regulatory elements to their target genes. Here we develop HiCPlus, a computational approach based on deep convolutional neural network, to infer high-resolution Hi-C interaction matrices from low-resolution Hi-C data. We demonstrate that HiCPlus can impute interaction matrices highly similar to the original ones, while only using 1/16 of the original sequencing reads. We show that the models learned from one cell type can be applied to make predictions in other cell or tissue types. Our work not only provides a computational framework to enhance Hi-C data resolution but also reveals features underlying the formation of 3D chromatin interactions.

  19. 3D Convolutional Neural Networks for Crop Classification with Multi-Temporal Remote Sensing Images

    Directory of Open Access Journals (Sweden)

    Shunping Ji

    2018-01-01

    Full Text Available This study describes a novel three-dimensional (3D convolutional neural networks (CNN based method that automatically classifies crops from spatio-temporal remote sensing images. First, 3D kernel is designed according to the structure of multi-spectral multi-temporal remote sensing data. Secondly, the 3D CNN framework with fine-tuned parameters is designed for training 3D crop samples and learning spatio-temporal discriminative representations, with the full crop growth cycles being preserved. In addition, we introduce an active learning strategy to the CNN model to improve labelling accuracy up to a required threshold with the most efficiency. Finally, experiments are carried out to test the advantage of the 3D CNN, in comparison to the two-dimensional (2D CNN and other conventional methods. Our experiments show that the 3D CNN is especially suitable in characterizing the dynamics of crop growth and outperformed the other mainstream methods.

  20. Subject independent facial expression recognition with robust face detection using a convolutional neural network.

    Science.gov (United States)

    Matsugu, Masakazu; Mori, Katsuhiko; Mitari, Yusuke; Kaneda, Yuji

    2003-01-01

    Reliable detection of ordinary facial expressions (e.g. smile) despite the variability among individuals as well as face appearance is an important step toward the realization of perceptual user interface with autonomous perception of persons. We describe a rule-based algorithm for robust facial expression recognition combined with robust face detection using a convolutional neural network. In this study, we address the problem of subject independence as well as translation, rotation, and scale invariance in the recognition of facial expression. The result shows reliable detection of smiles with recognition rate of 97.6% for 5600 still images of more than 10 subjects. The proposed algorithm demonstrated the ability to discriminate smiling from talking based on the saliency score obtained from voting visual cues. To the best of our knowledge, it is the first facial expression recognition model with the property of subject independence combined with robustness to variability in facial appearance.

  1. Imaging in scattering media using correlation image sensors and sparse convolutional coding

    KAUST Repository

    Heide, Felix; Xiao, Lei; Kolb, Andreas; Hullin, Matthias B.; Heidrich, Wolfgang

    2014-01-01

    Correlation image sensors have recently become popular low-cost devices for time-of-flight, or range cameras. They usually operate under the assumption of a single light path contributing to each pixel. We show that a more thorough analysis of the sensor data from correlation sensors can be used can be used to analyze the light transport in much more complex environments, including applications for imaging through scattering and turbid media. The key of our method is a new convolutional sparse coding approach for recovering transient (light-in-flight) images from correlation image sensors. This approach is enabled by an analysis of sparsity in complex transient images, and the derivation of a new physically-motivated model for transient images with drastically improved sparsity.

  2. Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties

    Science.gov (United States)

    Xie, Tian; Grossman, Jeffrey C.

    2018-04-01

    The use of machine learning methods for accelerating the design of crystalline materials usually requires manually constructed feature vectors or complex transformation of atom coordinates to input the crystal structure, which either constrains the model to certain crystal types or makes it difficult to provide chemical insights. Here, we develop a crystal graph convolutional neural networks framework to directly learn material properties from the connection of atoms in the crystal, providing a universal and interpretable representation of crystalline materials. Our method provides a highly accurate prediction of density functional theory calculated properties for eight different properties of crystals with various structure types and compositions after being trained with 1 04 data points. Further, our framework is interpretable because one can extract the contributions from local chemical environments to global properties. Using an example of perovskites, we show how this information can be utilized to discover empirical rules for materials design.

  3. Intelligent Image Recognition System for Marine Fouling Using Softmax Transfer Learning and Deep Convolutional Neural Networks

    Directory of Open Access Journals (Sweden)

    C. S. Chin

    2017-01-01

    Full Text Available The control of biofouling on marine vessels is challenging and costly. Early detection before hull performance is significantly affected is desirable, especially if “grooming” is an option. Here, a system is described to detect marine fouling at an early stage of development. In this study, an image of fouling can be transferred wirelessly via a mobile network for analysis. The proposed system utilizes transfer learning and deep convolutional neural network (CNN to perform image recognition on the fouling image by classifying the detected fouling species and the density of fouling on the surface. Transfer learning using Google’s Inception V3 model with Softmax at last layer was carried out on a fouling database of 10 categories and 1825 images. Experimental results gave acceptable accuracies for fouling detection and recognition.

  4. Imaging in scattering media using correlation image sensors and sparse convolutional coding

    KAUST Repository

    Heide, Felix

    2014-10-17

    Correlation image sensors have recently become popular low-cost devices for time-of-flight, or range cameras. They usually operate under the assumption of a single light path contributing to each pixel. We show that a more thorough analysis of the sensor data from correlation sensors can be used can be used to analyze the light transport in much more complex environments, including applications for imaging through scattering and turbid media. The key of our method is a new convolutional sparse coding approach for recovering transient (light-in-flight) images from correlation image sensors. This approach is enabled by an analysis of sparsity in complex transient images, and the derivation of a new physically-motivated model for transient images with drastically improved sparsity.

  5. Classification of time-series images using deep convolutional neural networks

    Science.gov (United States)

    Hatami, Nima; Gavet, Yann; Debayle, Johan

    2018-04-01

    Convolutional Neural Networks (CNN) has achieved a great success in image recognition task by automatically learning a hierarchical feature representation from raw data. While the majority of Time-Series Classification (TSC) literature is focused on 1D signals, this paper uses Recurrence Plots (RP) to transform time-series into 2D texture images and then take advantage of the deep CNN classifier. Image representation of time-series introduces different feature types that are not available for 1D signals, and therefore TSC can be treated as texture image recognition task. CNN model also allows learning different levels of representations together with a classifier, jointly and automatically. Therefore, using RP and CNN in a unified framework is expected to boost the recognition rate of TSC. Experimental results on the UCR time-series classification archive demonstrate competitive accuracy of the proposed approach, compared not only to the existing deep architectures, but also to the state-of-the art TSC algorithms.

  6. Cloud Detection by Fusing Multi-Scale Convolutional Features

    Science.gov (United States)

    Li, Zhiwei; Shen, Huanfeng; Wei, Yancong; Cheng, Qing; Yuan, Qiangqiang

    2018-04-01

    Clouds detection is an important pre-processing step for accurate application of optical satellite imagery. Recent studies indicate that deep learning achieves best performance in image segmentation tasks. Aiming at boosting the accuracy of cloud detection for multispectral imagery, especially for those that contain only visible and near infrared bands, in this paper, we proposed a deep learning based cloud detection method termed MSCN (multi-scale cloud net), which segments cloud by fusing multi-scale convolutional features. MSCN was trained on a global cloud cover validation collection, and was tested in more than ten types of optical images with different resolution. Experiment results show that MSCN has obvious advantages over the traditional multi-feature combined cloud detection method in accuracy, especially when in snow and other areas covered by bright non-cloud objects. Besides, MSCN produced more detailed cloud masks than the compared deep cloud detection convolution network. The effectiveness of MSCN make it promising for practical application in multiple kinds of optical imagery.

  7. Real-Time Video Convolutional Face Finder on Embedded Platforms

    Directory of Open Access Journals (Sweden)

    Mamalet Franck

    2007-01-01

    Full Text Available A high-level optimization methodology is applied for implementing the well-known convolutional face finder (CFF algorithm for real-time applications on mobile phones, such as teleconferencing, advanced user interfaces, image indexing, and security access control. CFF is based on a feature extraction and classification technique which consists of a pipeline of convolutions and subsampling operations. The design of embedded systems requires a good trade-off between performance and code size due to the limited amount of available resources. The followed methodology copes with the main drawbacks of the original implementation of CFF such as floating-point computation and memory allocation, in order to allow parallelism exploitation and perform algorithm optimizations. Experimental results show that our embedded face detection system can accurately locate faces with less computational load and memory cost. It runs on a 275 MHz Starcore DSP at 35 QCIF images/s with state-of-the-art detection rates and very low false alarm rates.

  8. Enhancing neutron beam production with a convoluted moderator

    Energy Technology Data Exchange (ETDEWEB)

    Iverson, E.B., E-mail: iversoneb@ornl.gov [Spallation Neutron Source, Oak Ridge National Laboratory, Oak Ridge, TN 37831 (United States); Baxter, D.V. [Center for the Exploration of Energy and Matter, Indiana University, Bloomington, IN 47408 (United States); Muhrer, G. [Lujan Neutron Scattering Center, Los Alamos National Laboratory, P.O. Box 1663, Los Alamos, NM 87545 (United States); Ansell, S.; Dalgliesh, R. [ISIS Facility, Rutherford Appleton Laboratory, Chilton (United Kingdom); Gallmeier, F.X. [Spallation Neutron Source, Oak Ridge National Laboratory, Oak Ridge, TN 37831 (United States); Kaiser, H. [Center for the Exploration of Energy and Matter, Indiana University, Bloomington, IN 47408 (United States); Lu, W. [Spallation Neutron Source, Oak Ridge National Laboratory, Oak Ridge, TN 37831 (United States)

    2014-10-21

    We describe a new concept for a neutron moderating assembly resulting in the more efficient production of slow neutron beams. The Convoluted Moderator, a heterogeneous stack of interleaved moderating material and nearly transparent single-crystal spacers, is a directionally enhanced neutron beam source, improving beam emission over an angular range comparable to the range accepted by neutron beam lines and guides. We have demonstrated gains of 50% in slow neutron intensity for a given fast neutron production rate while simultaneously reducing the wavelength-dependent emission time dispersion by 25%, both coming from a geometric effect in which the neutron beam lines view a large surface area of moderating material in a relatively small volume. Additionally, we have confirmed a Bragg-enhancement effect arising from coherent scattering within the single-crystal spacers. We have not observed hypothesized refractive effects leading to additional gains at long wavelength. In addition to confirmation of the validity of the Convoluted Moderator concept, our measurements provide a series of benchmark experiments suitable for developing simulation and analysis techniques for practical optimization and eventual implementation at slow neutron source facilities.

  9. Photon beam convolution using polyenergetic energy deposition kernels

    International Nuclear Information System (INIS)

    Hoban, P.W.; Murray, D.C.; Round, W.H.

    1994-01-01

    In photon beam convolution calculations where polyenergetic energy deposition kernels (EDKs) are used, the primary photon energy spectrum should be correctly accounted for in Monte Carlo generation of EDKs. This requires the probability of interaction, determined by the linear attenuation coefficient, μ, to be taken into account when primary photon interactions are forced to occur at the EDK origin. The use of primary and scattered EDKs generated with a fixed photon spectrum can give rise to an error in the dose calculation due to neglecting the effects of beam hardening with depth. The proportion of primary photon energy that is transferred to secondary electrons increases with depth of interaction, due to the increase in the ratio μ ab /μ as the beam hardens. Convolution depth-dose curves calculated using polyenergetic EDKs generated for the primary photon spectra which exist at depths of 0, 20 and 40 cm in water, show a fall-off which is too steep when compared with EGS4 Monte Carlo results. A beam hardening correction factor applied to primary and scattered 0 cm EDKs, based on the ratio of kerma to terma at each depth, gives primary, scattered and total dose in good agreement with Monte Carlo results. (Author)

  10. Spatiotemporal Recurrent Convolutional Networks for Traffic Prediction in Transportation Networks

    Directory of Open Access Journals (Sweden)

    Haiyang Yu

    2017-06-01

    Full Text Available Predicting large-scale transportation network traffic has become an important and challenging topic in recent decades. Inspired by the domain knowledge of motion prediction, in which the future motion of an object can be predicted based on previous scenes, we propose a network grid representation method that can retain the fine-scale structure of a transportation network. Network-wide traffic speeds are converted into a series of static images and input into a novel deep architecture, namely, spatiotemporal recurrent convolutional networks (SRCNs, for traffic forecasting. The proposed SRCNs inherit the advantages of deep convolutional neural networks (DCNNs and long short-term memory (LSTM neural networks. The spatial dependencies of network-wide traffic can be captured by DCNNs, and the temporal dynamics can be learned by LSTMs. An experiment on a Beijing transportation network with 278 links demonstrates that SRCNs outperform other deep learning-based algorithms in both short-term and long-term traffic prediction.

  11. Convolutional Neural Network for Histopathological Analysis of Osteosarcoma.

    Science.gov (United States)

    Mishra, Rashika; Daescu, Ovidiu; Leavey, Patrick; Rakheja, Dinesh; Sengupta, Anita

    2018-03-01

    Pathologists often deal with high complexity and sometimes disagreement over osteosarcoma tumor classification due to cellular heterogeneity in the dataset. Segmentation and classification of histology tissue in H&E stained tumor image datasets is a challenging task because of intra-class variations, inter-class similarity, crowded context, and noisy data. In recent years, deep learning approaches have led to encouraging results in breast cancer and prostate cancer analysis. In this article, we propose convolutional neural network (CNN) as a tool to improve efficiency and accuracy of osteosarcoma tumor classification into tumor classes (viable tumor, necrosis) versus nontumor. The proposed CNN architecture contains eight learned layers: three sets of stacked two convolutional layers interspersed with max pooling layers for feature extraction and two fully connected layers with data augmentation strategies to boost performance. The use of a neural network results in higher accuracy of average 92% for the classification. We compare the proposed architecture with three existing and proven CNN architectures for image classification: AlexNet, LeNet, and VGGNet. We also provide a pipeline to calculate percentage necrosis in a given whole slide image. We conclude that the use of neural networks can assure both high accuracy and efficiency in osteosarcoma classification.

  12. Real-Time Video Convolutional Face Finder on Embedded Platforms

    Directory of Open Access Journals (Sweden)

    Franck Mamalet

    2007-03-01

    Full Text Available A high-level optimization methodology is applied for implementing the well-known convolutional face finder (CFF algorithm for real-time applications on mobile phones, such as teleconferencing, advanced user interfaces, image indexing, and security access control. CFF is based on a feature extraction and classification technique which consists of a pipeline of convolutions and subsampling operations. The design of embedded systems requires a good trade-off between performance and code size due to the limited amount of available resources. The followed methodology copes with the main drawbacks of the original implementation of CFF such as floating-point computation and memory allocation, in order to allow parallelism exploitation and perform algorithm optimizations. Experimental results show that our embedded face detection system can accurately locate faces with less computational load and memory cost. It runs on a 275 MHz Starcore DSP at 35 QCIF images/s with state-of-the-art detection rates and very low false alarm rates.

  13. Digital image correlation based on a fast convolution strategy

    Science.gov (United States)

    Yuan, Yuan; Zhan, Qin; Xiong, Chunyang; Huang, Jianyong

    2017-10-01

    In recent years, the efficiency of digital image correlation (DIC) methods has attracted increasing attention because of its increasing importance for many engineering applications. Based on the classical affine optical flow (AOF) algorithm and the well-established inverse compositional Gauss-Newton algorithm, which is essentially a natural extension of the AOF algorithm under a nonlinear iterative framework, this paper develops a set of fast convolution-based DIC algorithms for high-efficiency subpixel image registration. Using a well-developed fast convolution technique, the set of algorithms establishes a series of global data tables (GDTs) over the digital images, which allows the reduction of the computational complexity of DIC significantly. Using the pre-calculated GDTs, the subpixel registration calculations can be implemented efficiently in a look-up-table fashion. Both numerical simulation and experimental verification indicate that the set of algorithms significantly enhances the computational efficiency of DIC, especially in the case of a dense data sampling for the digital images. Because the GDTs need to be computed only once, the algorithms are also suitable for efficiently coping with image sequences that record the time-varying dynamics of specimen deformations.

  14. Convolutional neural network features based change detection in satellite images

    Science.gov (United States)

    Mohammed El Amin, Arabi; Liu, Qingjie; Wang, Yunhong

    2016-07-01

    With the popular use of high resolution remote sensing (HRRS) satellite images, a huge research efforts have been placed on change detection (CD) problem. An effective feature selection method can significantly boost the final result. While hand-designed features have proven difficulties to design features that effectively capture high and mid-level representations, the recent developments in machine learning (Deep Learning) omit this problem by learning hierarchical representation in an unsupervised manner directly from data without human intervention. In this letter, we propose approaching the change detection problem from a feature learning perspective. A novel deep Convolutional Neural Networks (CNN) features based HR satellite images change detection method is proposed. The main guideline is to produce a change detection map directly from two images using a pretrained CNN. This method can omit the limited performance of hand-crafted features. Firstly, CNN features are extracted through different convolutional layers. Then, a concatenation step is evaluated after an normalization step, resulting in a unique higher dimensional feature map. Finally, a change map was computed using pixel-wise Euclidean distance. Our method has been validated on real bitemporal HRRS satellite images according to qualitative and quantitative analyses. The results obtained confirm the interest of the proposed method.

  15. Classification of stroke disease using convolutional neural network

    Science.gov (United States)

    Marbun, J. T.; Seniman; Andayani, U.

    2018-03-01

    Stroke is a condition that occurs when the blood supply stop flowing to the brain because of a blockage or a broken blood vessel. A symptoms that happen when experiencing stroke, some of them is a dropped consciousness, disrupted vision and paralyzed body. The general examination is being done to get a picture of the brain part that have stroke using Computerized Tomography (CT) Scan. The image produced from CT will be manually checked and need a proper lighting by doctor to get a type of stroke. That is why it needs a method to classify stroke from CT image automatically. A method proposed in this research is Convolutional Neural Network. CT image of the brain is used as the input for image processing. The stage before classification are image processing (Grayscaling, Scaling, Contrast Limited Adaptive Histogram Equalization, then the image being classified with Convolutional Neural Network. The result then showed that the method significantly conducted was able to be used as a tool to classify stroke disease in order to distinguish the type of stroke from CT image.

  16. Image Classification Based on Convolutional Denoising Sparse Autoencoder

    Directory of Open Access Journals (Sweden)

    Shuangshuang Chen

    2017-01-01

    Full Text Available Image classification aims to group images into corresponding semantic categories. Due to the difficulties of interclass similarity and intraclass variability, it is a challenging issue in computer vision. In this paper, an unsupervised feature learning approach called convolutional denoising sparse autoencoder (CDSAE is proposed based on the theory of visual attention mechanism and deep learning methods. Firstly, saliency detection method is utilized to get training samples for unsupervised feature learning. Next, these samples are sent to the denoising sparse autoencoder (DSAE, followed by convolutional layer and local contrast normalization layer. Generally, prior in a specific task is helpful for the task solution. Therefore, a new pooling strategy—spatial pyramid pooling (SPP fused with center-bias prior—is introduced into our approach. Experimental results on the common two image datasets (STL-10 and CIFAR-10 demonstrate that our approach is effective in image classification. They also demonstrate that none of these three components: local contrast normalization, SPP fused with center-prior, and l2 vector normalization can be excluded from our proposed approach. They jointly improve image representation and classification performance.

  17. Weed Growth Stage Estimator Using Deep Convolutional Neural Networks.

    Science.gov (United States)

    Teimouri, Nima; Dyrmann, Mads; Nielsen, Per Rydahl; Mathiassen, Solvejg Kopp; Somerville, Gayle J; Jørgensen, Rasmus Nyholm

    2018-05-16

    This study outlines a new method of automatically estimating weed species and growth stages (from cotyledon until eight leaves are visible) of in situ images covering 18 weed species or families. Images of weeds growing within a variety of crops were gathered across variable environmental conditions with regards to soil types, resolution and light settings. Then, 9649 of these images were used for training the computer, which automatically divided the weeds into nine growth classes. The performance of this proposed convolutional neural network approach was evaluated on a further set of 2516 images, which also varied in term of crop, soil type, image resolution and light conditions. The overall performance of this approach achieved a maximum accuracy of 78% for identifying Polygonum spp. and a minimum accuracy of 46% for blackgrass. In addition, it achieved an average 70% accuracy rate in estimating the number of leaves and 96% accuracy when accepting a deviation of two leaves. These results show that this new method of using deep convolutional neural networks has a relatively high ability to estimate early growth stages across a wide variety of weed species.

  18. sEMG-Based Gesture Recognition with Convolution Neural Networks

    Directory of Open Access Journals (Sweden)

    Zhen Ding

    2018-06-01

    Full Text Available The traditional classification methods for limb motion recognition based on sEMG have been deeply researched and shown promising results. However, information loss during feature extraction reduces the recognition accuracy. To obtain higher accuracy, the deep learning method was introduced. In this paper, we propose a parallel multiple-scale convolution architecture. Compared with the state-of-art methods, the proposed architecture fully considers the characteristics of the sEMG signal. Larger sizes of kernel filter than commonly used in other CNN-based hand recognition methods are adopted. Meanwhile, the characteristics of the sEMG signal, that is, muscle independence, is considered when designing the architecture. All the classification methods were evaluated on the NinaPro database. The results show that the proposed architecture has the highest recognition accuracy. Furthermore, the results indicate that parallel multiple-scale convolution architecture with larger size of kernel filter and considering muscle independence can significantly increase the classification accuracy.

  19. Development of a morphological convolution operator for bearing fault detection

    Science.gov (United States)

    Li, Yifan; Liang, Xihui; Liu, Weiwei; Wang, Yan

    2018-05-01

    This paper presents a novel signal processing scheme, namely morphological convolution operator (MCO) lifted morphological undecimated wavelet (MUDW), for rolling element bearing fault detection. In this scheme, a MCO is first designed to fully utilize the advantage of the closing & opening gradient operator and the closing-opening & opening-closing gradient operator for feature extraction as well as the merit of excellent denoising characteristics of the convolution operator. The MCO is then introduced into MUDW for the purpose of improving the fault detection ability of the reported MUDWs. Experimental vibration signals collected from a train wheelset test rig and the bearing data center of Case Western Reserve University are employed to evaluate the effectiveness of the proposed MCO lifted MUDW on fault detection of rolling element bearings. The results show that the proposed approach has a superior performance in extracting fault features of defective rolling element bearings. In addition, comparisons are performed between two reported MUDWs and the proposed MCO lifted MUDW. The MCO lifted MUDW outperforms both of them in detection of outer race faults and inner race faults of rolling element bearings.

  20. Fluence-convolution broad-beam (FCBB) dose calculation

    Energy Technology Data Exchange (ETDEWEB)

    Lu Weiguo; Chen Mingli, E-mail: wlu@tomotherapy.co [TomoTherapy Inc., 1240 Deming Way, Madison, WI 53717 (United States)

    2010-12-07

    IMRT optimization requires a fast yet relatively accurate algorithm to calculate the iteration dose with small memory demand. In this paper, we present a dose calculation algorithm that approaches these goals. By decomposing the infinitesimal pencil beam (IPB) kernel into the central axis (CAX) component and lateral spread function (LSF) and taking the beam's eye view (BEV), we established a non-voxel and non-beamlet-based dose calculation formula. Both LSF and CAX are determined by a commissioning procedure using the collapsed-cone convolution/superposition (CCCS) method as the standard dose engine. The proposed dose calculation involves a 2D convolution of a fluence map with LSF followed by ray tracing based on the CAX lookup table with radiological distance and divergence correction, resulting in complexity of O(N{sup 3}) both spatially and temporally. This simple algorithm is orders of magnitude faster than the CCCS method. Without pre-calculation of beamlets, its implementation is also orders of magnitude smaller than the conventional voxel-based beamlet-superposition (VBS) approach. We compared the presented algorithm with the CCCS method using simulated and clinical cases. The agreement was generally within 3% for a homogeneous phantom and 5% for heterogeneous and clinical cases. Combined with the 'adaptive full dose correction', the algorithm is well suitable for calculating the iteration dose during IMRT optimization.

  1. Deep multi-scale convolutional neural network for hyperspectral image classification

    Science.gov (United States)

    Zhang, Feng-zhe; Yang, Xia

    2018-04-01

    In this paper, we proposed a multi-scale convolutional neural network for hyperspectral image classification task. Firstly, compared with conventional convolution, we utilize multi-scale convolutions, which possess larger respective fields, to extract spectral features of hyperspectral image. We design a deep neural network with a multi-scale convolution layer which contains 3 different convolution kernel sizes. Secondly, to avoid overfitting of deep neural network, dropout is utilized, which randomly sleeps neurons, contributing to improve the classification accuracy a bit. In addition, new skills like ReLU in deep learning is utilized in this paper. We conduct experiments on University of Pavia and Salinas datasets, and obtained better classification accuracy compared with other methods.

  2. Object Detection Based on Fast/Faster RCNN Employing Fully Convolutional Architectures

    Directory of Open Access Journals (Sweden)

    Yun Ren

    2018-01-01

    Full Text Available Modern object detectors always include two major parts: a feature extractor and a feature classifier as same as traditional object detectors. The deeper and wider convolutional architectures are adopted as the feature extractor at present. However, many notable object detection systems such as Fast/Faster RCNN only consider simple fully connected layers as the feature classifier. In this paper, we declare that it is beneficial for the detection performance to elaboratively design deep convolutional networks (ConvNets of various depths for feature classification, especially using the fully convolutional architectures. In addition, this paper also demonstrates how to employ the fully convolutional architectures in the Fast/Faster RCNN. Experimental results show that a classifier based on convolutional layer is more effective for object detection than that based on fully connected layer and that the better detection performance can be achieved by employing deeper ConvNets as the feature classifier.

  3. Constructing fine-granularity functional brain network atlases via deep convolutional autoencoder.

    Science.gov (United States)

    Zhao, Yu; Dong, Qinglin; Chen, Hanbo; Iraji, Armin; Li, Yujie; Makkie, Milad; Kou, Zhifeng; Liu, Tianming

    2017-12-01

    State-of-the-art functional brain network reconstruction methods such as independent component analysis (ICA) or sparse coding of whole-brain fMRI data can effectively infer many thousands of volumetric brain network maps from a large number of human brains. However, due to the variability of individual brain networks and the large scale of such networks needed for statistically meaningful group-level analysis, it is still a challenging and open problem to derive group-wise common networks as network atlases. Inspired by the superior spatial pattern description ability of the deep convolutional neural networks (CNNs), a novel deep 3D convolutional autoencoder (CAE) network is designed here to extract spatial brain network features effectively, based on which an Apache Spark enabled computational framework is developed for fast clustering of larger number of network maps into fine-granularity atlases. To evaluate this framework, 10 resting state networks (RSNs) were manually labeled from the sparsely decomposed networks of Human Connectome Project (HCP) fMRI data and 5275 network training samples were obtained, in total. Then the deep CAE models are trained by these functional networks' spatial maps, and the learned features are used to refine the original 10 RSNs into 17 network atlases that possess fine-granularity functional network patterns. Interestingly, it turned out that some manually mislabeled outliers in training networks can be corrected by the deep CAE derived features. More importantly, fine granularities of networks can be identified and they reveal unique network patterns specific to different brain task states. By further applying this method to a dataset of mild traumatic brain injury study, it shows that the technique can effectively identify abnormal small networks in brain injury patients in comparison with controls. In general, our work presents a promising deep learning and big data analysis solution for modeling functional connectomes, with

  4. Three-Dimensional Terahertz Coded-Aperture Imaging Based on Matched Filtering and Convolutional Neural Network.

    Science.gov (United States)

    Chen, Shuo; Luo, Chenggao; Wang, Hongqiang; Deng, Bin; Cheng, Yongqiang; Zhuang, Zhaowen

    2018-04-26

    As a promising radar imaging technique, terahertz coded-aperture imaging (TCAI) can achieve high-resolution, forward-looking, and staring imaging by producing spatiotemporal independent signals with coded apertures. However, there are still two problems in three-dimensional (3D) TCAI. Firstly, the large-scale reference-signal matrix based on meshing the 3D imaging area creates a heavy computational burden, thus leading to unsatisfactory efficiency. Secondly, it is difficult to resolve the target under low signal-to-noise ratio (SNR). In this paper, we propose a 3D imaging method based on matched filtering (MF) and convolutional neural network (CNN), which can reduce the computational burden and achieve high-resolution imaging for low SNR targets. In terms of the frequency-hopping (FH) signal, the original echo is processed with MF. By extracting the processed echo in different spike pulses separately, targets in different imaging planes are reconstructed simultaneously to decompose the global computational complexity, and then are synthesized together to reconstruct the 3D target. Based on the conventional TCAI model, we deduce and build a new TCAI model based on MF. Furthermore, the convolutional neural network (CNN) is designed to teach the MF-TCAI how to reconstruct the low SNR target better. The experimental results demonstrate that the MF-TCAI achieves impressive performance on imaging ability and efficiency under low SNR. Moreover, the MF-TCAI has learned to better resolve the low-SNR 3D target with the help of CNN. In summary, the proposed 3D TCAI can achieve: (1) low-SNR high-resolution imaging by using MF; (2) efficient 3D imaging by downsizing the large-scale reference-signal matrix; and (3) intelligent imaging with CNN. Therefore, the TCAI based on MF and CNN has great potential in applications such as security screening, nondestructive detection, medical diagnosis, etc.

  5. Deep Convolutional Neural Networks for Multi-Modality Isointense Infant Brain Image Segmentation

    Science.gov (United States)

    Zhang, Wenlu; Li, Rongjian; Deng, Houtao; Wang, Li; Lin, Weili; Ji, Shuiwang; Shen, Dinggang

    2015-01-01

    The segmentation of infant brain tissue images into white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) plays an important role in studying early brain development in health and disease. In the isointense stage (approximately 6–8 months of age), WM and GM exhibit similar levels of intensity in both T1 and T2 MR images, making the tissue segmentation very challenging. Only a small number of existing methods have been designed for tissue segmentation in this isointense stage; however, they only used a single T1 or T2 images, or the combination of T1 and T2 images. In this paper, we propose to use deep convolutional neural networks (CNNs) for segmenting isointense stage brain tissues using multi-modality MR images. CNNs are a type of deep models in which trainable filters and local neighborhood pooling operations are applied alternatingly on the raw input images, resulting in a hierarchy of increasingly complex features. Specifically, we used multimodality information from T1, T2, and fractional anisotropy (FA) images as inputs and then generated the segmentation maps as outputs. The multiple intermediate layers applied convolution, pooling, normalization, and other operations to capture the highly nonlinear mappings between inputs and outputs. We compared the performance of our approach with that of the commonly used segmentation methods on a set of manually segmented isointense stage brain images. Results showed that our proposed model significantly outperformed prior methods on infant brain tissue segmentation. In addition, our results indicated that integration of multi-modality images led to significant performance improvement. PMID:25562829

  6. Introducing single-crystal scattering and optical potentials into MCNPX: Predicting neutron emission from a convoluted moderator

    Energy Technology Data Exchange (ETDEWEB)

    Gallmeier, F.X., E-mail: gallmeierfz@ornl.gov [Spallation Neutron Source, Oak Ridge National Laboratory, Oak Ridge, TN 37831 (United States); Iverson, E.B.; Lu, W. [Spallation Neutron Source, Oak Ridge National Laboratory, Oak Ridge, TN 37831 (United States); Baxter, D.V. [Center for the Exploration of Energy and Matter, Indiana University, Bloomington, IN 47408 (United States); Muhrer, G.; Ansell, S. [European Spallation Source, ESS AB, Lund (Sweden)

    2016-04-01

    Neutron transport simulation codes are indispensable tools for the design and construction of modern neutron scattering facilities and instrumentation. Recently, it has become increasingly clear that some neutron instrumentation has started to exploit physics that is not well-modeled by the existing codes. In particular, the transport of neutrons through single crystals and across interfaces in MCNP(X), Geant4, and other codes ignores scattering from oriented crystals and refractive effects, and yet these are essential phenomena for the performance of monochromators and ultra-cold neutron transport respectively (to mention but two examples). In light of these developments, we have extended the MCNPX code to include a single-crystal neutron scattering model and neutron reflection/refraction physics. We have also generated silicon scattering kernels for single crystals of definable orientation. As a first test of these new tools, we have chosen to model the recently developed convoluted moderator concept, in which a moderating material is interleaved with layers of perfect crystals to provide an exit path for neutrons moderated to energies below the crystal's Bragg cut–off from locations deep within the moderator. Studies of simple cylindrical convoluted moderator systems of 100 mm diameter and composed of polyethylene and single crystal silicon were performed with the upgraded MCNPX code and reproduced the magnitude of effects seen in experiments compared to homogeneous moderator systems. Applying different material properties for refraction and reflection, and by replacing the silicon in the models with voids, we show that the emission enhancements seen in recent experiments are primarily caused by the transparency of the silicon and void layers. Finally we simulated the convoluted moderator experiments described by Iverson et al. and found satisfactory agreement between the measurements and the simulations performed with the tools we have developed.

  7. Automated red blood cells extraction from holographic images using fully convolutional neural networks

    Science.gov (United States)

    Yi, Faliu; Moon, Inkyu; Javidi, Bahram

    2017-01-01

    In this paper, we present two models for automatically extracting red blood cells (RBCs) from RBCs holographic images based on a deep learning fully convolutional neural network (FCN) algorithm. The first model, called FCN-1, only uses the FCN algorithm to carry out RBCs prediction, whereas the second model, called FCN-2, combines the FCN approach with the marker-controlled watershed transform segmentation scheme to achieve RBCs extraction. Both models achieve good segmentation accuracy. In addition, the second model has much better performance in terms of cell separation than traditional segmentation methods. In the proposed methods, the RBCs phase images are first numerically reconstructed from RBCs holograms recorded with off-axis digital holographic microscopy. Then, some RBCs phase images are manually segmented and used as training data to fine-tune the FCN. Finally, each pixel in new input RBCs phase images is predicted into either foreground or background using the trained FCN models. The RBCs prediction result from the first model is the final segmentation result, whereas the result from the second model is used as the internal markers of the marker-controlled transform algorithm for further segmentation. Experimental results show that the given schemes can automatically extract RBCs from RBCs phase images and much better RBCs separation results are obtained when the FCN technique is combined with the marker-controlled watershed segmentation algorithm. PMID:29082078

  8. EMG-Based Estimation of Limb Movement Using Deep Learning With Recurrent Convolutional Neural Networks.

    Science.gov (United States)

    Xia, Peng; Hu, Jie; Peng, Yinghong

    2017-10-25

    A novel model based on deep learning is proposed to estimate kinematic information for myoelectric control from multi-channel electromyogram (EMG) signals. The neural information of limb movement is embedded in EMG signals that are influenced by all kinds of factors. In order to overcome the negative effects of variability in signals, the proposed model employs the deep architecture combining convolutional neural networks (CNNs) and recurrent neural networks (RNNs). The EMG signals are transformed to time-frequency frames as the input to the model. The limb movement is estimated by the model that is trained with the gradient descent and backpropagation procedure. We tested the model for simultaneous and proportional estimation of limb movement in eight healthy subjects and compared it with support vector regression (SVR) and CNNs on the same data set. The experimental studies show that the proposed model has higher estimation accuracy and better robustness with respect to time. The combination of CNNs and RNNs can improve the model performance compared with using CNNs alone. The model of deep architecture is promising in EMG decoding and optimization of network structures can increase the accuracy and robustness. © 2017 International Center for Artificial Organs and Transplantation and Wiley Periodicals, Inc.

  9. DanQ: a hybrid convolutional and recurrent deep neural network for quantifying the function of DNA sequences.

    Science.gov (United States)

    Quang, Daniel; Xie, Xiaohui

    2016-06-20

    Modeling the properties and functions of DNA sequences is an important, but challenging task in the broad field of genomics. This task is particularly difficult for non-coding DNA, the vast majority of which is still poorly understood in terms of function. A powerful predictive model for the function of non-coding DNA can have enormous benefit for both basic science and translational research because over 98% of the human genome is non-coding and 93% of disease-associated variants lie in these regions. To address this need, we propose DanQ, a novel hybrid convolutional and bi-directional long short-term memory recurrent neural network framework for predicting non-coding function de novo from sequence. In the DanQ model, the convolution layer captures regulatory motifs, while the recurrent layer captures long-term dependencies between the motifs in order to learn a regulatory 'grammar' to improve predictions. DanQ improves considerably upon other models across several metrics. For some regulatory markers, DanQ can achieve over a 50% relative improvement in the area under the precision-recall curve metric compared to related models. We have made the source code available at the github repository http://github.com/uci-cbcl/DanQ. © The Author(s) 2016. Published by Oxford University Press on behalf of Nucleic Acids Research.

  10. Predicting Response to Neoadjuvant Chemotherapy with PET Imaging Using Convolutional Neural Networks.

    Directory of Open Access Journals (Sweden)

    Petros-Pavlos Ypsilantis

    Full Text Available Imaging of cancer with 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET has become a standard component of diagnosis and staging in oncology, and is becoming more important as a quantitative monitor of individual response to therapy. In this article we investigate the challenging problem of predicting a patient's response to neoadjuvant chemotherapy from a single 18F-FDG PET scan taken prior to treatment. We take a "radiomics" approach whereby a large amount of quantitative features is automatically extracted from pretherapy PET images in order to build a comprehensive quantification of the tumor phenotype. While the dominant methodology relies on hand-crafted texture features, we explore the potential of automatically learning low- to high-level features directly from PET scans. We report on a study that compares the performance of two competing radiomics strategies: an approach based on state-of-the-art statistical classifiers using over 100 quantitative imaging descriptors, including texture features as well as standardized uptake values, and a convolutional neural network, 3S-CNN, trained directly from PET scans by taking sets of adjacent intra-tumor slices. Our experimental results, based on a sample of 107 patients with esophageal cancer, provide initial evidence that convolutional neural networks have the potential to extract PET imaging representations that are highly predictive of response to therapy. On this dataset, 3S-CNN achieves an average 80.7% sensitivity and 81.6% specificity in predicting non-responders, and outperforms other competing predictive models.

  11. High-Throughput Classification of Radiographs Using Deep Convolutional Neural Networks.

    Science.gov (United States)

    Rajkomar, Alvin; Lingam, Sneha; Taylor, Andrew G; Blum, Michael; Mongan, John

    2017-02-01

    The study aimed to determine if computer vision techniques rooted in deep learning can use a small set of radiographs to perform clinically relevant image classification with high fidelity. One thousand eight hundred eighty-five chest radiographs on 909 patients obtained between January 2013 and July 2015 at our institution were retrieved and anonymized. The source images were manually annotated as frontal or lateral and randomly divided into training, validation, and test sets. Training and validation sets were augmented to over 150,000 images using standard image manipulations. We then pre-trained a series of deep convolutional networks based on the open-source GoogLeNet with various transformations of the open-source ImageNet (non-radiology) images. These trained networks were then fine-tuned using the original and augmented radiology images. The model with highest validation accuracy was applied to our institutional test set and a publicly available set. Accuracy was assessed by using the Youden Index to set a binary cutoff for frontal or lateral classification. This retrospective study was IRB approved prior to initiation. A network pre-trained on 1.2 million greyscale ImageNet images and fine-tuned on augmented radiographs was chosen. The binary classification method correctly classified 100 % (95 % CI 99.73-100 %) of both our test set and the publicly available images. Classification was rapid, at 38 images per second. A deep convolutional neural network created using non-radiological images, and an augmented set of radiographs is effective in highly accurate classification of chest radiograph view type and is a feasible, rapid method for high-throughput annotation.

  12. Ordinal convolutional neural networks for predicting RDoC positive valence psychiatric symptom severity scores.

    Science.gov (United States)

    Rios, Anthony; Kavuluru, Ramakanth

    2017-11-01

    The CEGS N-GRID 2016 Shared Task in Clinical Natural Language Processing (NLP) provided a set of 1000 neuropsychiatric notes to participants as part of a competition to predict psychiatric symptom severity scores. This paper summarizes our methods, results, and experiences based on our participation in the second track of the shared task. Classical methods of text classification usually fall into one of three problem types: binary, multi-class, and multi-label classification. In this effort, we study ordinal regression problems with text data where misclassifications are penalized differently based on how far apart the ground truth and model predictions are on the ordinal scale. Specifically, we present our entries (methods and results) in the N-GRID shared task in predicting research domain criteria (RDoC) positive valence ordinal symptom severity scores (absent, mild, moderate, and severe) from psychiatric notes. We propose a novel convolutional neural network (CNN) model designed to handle ordinal regression tasks on psychiatric notes. Broadly speaking, our model combines an ordinal loss function, a CNN, and conventional feature engineering (wide features) into a single model which is learned end-to-end. Given interpretability is an important concern with nonlinear models, we apply a recent approach called locally interpretable model-agnostic explanation (LIME) to identify important words that lead to instance specific predictions. Our best model entered into the shared task placed third among 24 teams and scored a macro mean absolute error (MMAE) based normalized score (100·(1-MMAE)) of 83.86. Since the competition, we improved our score (using basic ensembling) to 85.55, comparable with the winning shared task entry. Applying LIME to model predictions, we demonstrate the feasibility of instance specific prediction interpretation by identifying words that led to a particular decision. In this paper, we present a method that successfully uses wide features and

  13. Training strategy for convolutional neural networks in pedestrian gender classification

    Science.gov (United States)

    Ng, Choon-Boon; Tay, Yong-Haur; Goi, Bok-Min

    2017-06-01

    In this work, we studied a strategy for training a convolutional neural network in pedestrian gender classification with limited amount of labeled training data. Unsupervised learning by k-means clustering on pedestrian images was used to learn the filters to initialize the first layer of the network. As a form of pre-training, supervised learning for the related task of pedestrian classification was performed. Finally, the network was fine-tuned for gender classification. We found that this strategy improved the network's generalization ability in gender classification, achieving better test results when compared to random weights initialization and slightly more beneficial than merely initializing the first layer filters by unsupervised learning. This shows that unsupervised learning followed by pre-training with pedestrian images is an effective strategy to learn useful features for pedestrian gender classification.

  14. High Order Tensor Formulation for Convolutional Sparse Coding

    KAUST Repository

    Bibi, Adel Aamer

    2017-12-25

    Convolutional sparse coding (CSC) has gained attention for its successful role as a reconstruction and a classification tool in the computer vision and machine learning community. Current CSC methods can only reconstruct singlefeature 2D images independently. However, learning multidimensional dictionaries and sparse codes for the reconstruction of multi-dimensional data is very important, as it examines correlations among all the data jointly. This provides more capacity for the learned dictionaries to better reconstruct data. In this paper, we propose a generic and novel formulation for the CSC problem that can handle an arbitrary order tensor of data. Backed with experimental results, our proposed formulation can not only tackle applications that are not possible with standard CSC solvers, including colored video reconstruction (5D- tensors), but it also performs favorably in reconstruction with much fewer parameters as compared to naive extensions of standard CSC to multiple features/channels.

  15. CONEDEP: COnvolutional Neural network based Earthquake DEtection and Phase Picking

    Science.gov (United States)

    Zhou, Y.; Huang, Y.; Yue, H.; Zhou, S.; An, S.; Yun, N.

    2017-12-01

    We developed an automatic local earthquake detection and phase picking algorithm based on Fully Convolutional Neural network (FCN). The FCN algorithm detects and segments certain features (phases) in 3 component seismograms to realize efficient picking. We use STA/LTA algorithm and template matching algorithm to construct the training set from seismograms recorded 1 month before and after the Wenchuan earthquake. Precise P and S phases are identified and labeled to construct the training set. Noise data are produced by combining back-ground noise and artificial synthetic noise to form the equivalent scale of noise set as the signal set. Training is performed on GPUs to achieve efficient convergence. Our algorithm has significantly improved performance in terms of the detection rate and precision in comparison with STA/LTA and template matching algorithms.

  16. Computational optical tomography using 3-D deep convolutional neural networks

    Science.gov (United States)

    Nguyen, Thanh; Bui, Vy; Nehmetallah, George

    2018-04-01

    Deep convolutional neural networks (DCNNs) offer a promising performance for many image processing areas, such as super-resolution, deconvolution, image classification, denoising, and segmentation, with outstanding results. Here, we develop for the first time, to our knowledge, a method to perform 3-D computational optical tomography using 3-D DCNN. A simulated 3-D phantom dataset was first constructed and converted to a dataset of phase objects imaged on a spatial light modulator. For each phase image in the dataset, the corresponding diffracted intensity image was experimentally recorded on a CCD. We then experimentally demonstrate the ability of the developed 3-D DCNN algorithm to solve the inverse problem by reconstructing the 3-D index of refraction distributions of test phantoms from the dataset from their corresponding diffraction patterns.

  17. Drug-Drug Interaction Extraction via Convolutional Neural Networks

    Directory of Open Access Journals (Sweden)

    Shengyu Liu

    2016-01-01

    Full Text Available Drug-drug interaction (DDI extraction as a typical relation extraction task in natural language processing (NLP has always attracted great attention. Most state-of-the-art DDI extraction systems are based on support vector machines (SVM with a large number of manually defined features. Recently, convolutional neural networks (CNN, a robust machine learning method which almost does not need manually defined features, has exhibited great potential for many NLP tasks. It is worth employing CNN for DDI extraction, which has never been investigated. We proposed a CNN-based method for DDI extraction. Experiments conducted on the 2013 DDIExtraction challenge corpus demonstrate that CNN is a good choice for DDI extraction. The CNN-based DDI extraction method achieves an F-score of 69.75%, which outperforms the existing best performing method by 2.75%.

  18. Truncation Depth Rule-of-Thumb for Convolutional Codes

    Science.gov (United States)

    Moision, Bruce

    2009-01-01

    In this innovation, it is shown that a commonly used rule of thumb (that the truncation depth of a convolutional code should be five times the memory length, m, of the code) is accurate only for rate 1/2 codes. In fact, the truncation depth should be 2.5 m/(1 - r), where r is the code rate. The accuracy of this new rule is demonstrated by tabulating the distance properties of a large set of known codes. This new rule was derived by bounding the losses due to truncation as a function of the code rate. With regard to particular codes, a good indicator of the required truncation depth is the path length at which all paths that diverge from a particular path have accumulated the minimum distance of the code. It is shown that the new rule of thumb provides an accurate prediction of this depth for codes of varying rates.

  19. Finding Neutrinos in LArTPCs using Convolutional Neural Networks

    Science.gov (United States)

    Wongjirad, Taritree

    2017-09-01

    Deep learning algorithms, which have emerged over the last decade, are opening up new ways to analyze data for many particle physics experiments. MicroBooNE, which is a neutrino experiment at Fermilab, has been exploring the use of such algorithms, in particular, convolutional neural networks (CNNS). CNNs are the state-of-the-art method for a large class of problems involving the analysis of images. This makes CNNs an attractive approach for MicroBooNE, whose detector, a liquid argon time projection chamber (LArTPC), produces high-resolution images of particle interactions. In this talk, I will discuss the ways CNNs can be applied to tasks like neutrino interaction detection and particle identification in MicroBooNE and LArTPCs.

  20. Radio frequency interference mitigation using deep convolutional neural networks

    Science.gov (United States)

    Akeret, J.; Chang, C.; Lucchi, A.; Refregier, A.

    2017-01-01

    We propose a novel approach for mitigating radio frequency interference (RFI) signals in radio data using the latest advances in deep learning. We employ a special type of Convolutional Neural Network, the U-Net, that enables the classification of clean signal and RFI signatures in 2D time-ordered data acquired from a radio telescope. We train and assess the performance of this network using the HIDE &SEEK radio data simulation and processing packages, as well as early Science Verification data acquired with the 7m single-dish telescope at the Bleien Observatory. We find that our U-Net implementation is showing competitive accuracy to classical RFI mitigation algorithms such as SEEK's SUMTHRESHOLD implementation. We publish our U-Net software package on GitHub under GPLv3 license.

  1. Forecasting Flare Activity Using Deep Convolutional Neural Networks

    Science.gov (United States)

    Hernandez, T.

    2017-12-01

    Current operational flare forecasting relies on human morphological analysis of active regions and the persistence of solar flare activity through time (i.e. that the Sun will continue to do what it is doing right now: flaring or remaining calm). In this talk we present the results of applying deep Convolutional Neural Networks (CNNs) to the problem of solar flare forecasting. CNNs operate by training a set of tunable spatial filters that, in combination with neural layer interconnectivity, allow CNNs to automatically identify significant spatial structures predictive for classification and regression problems. We will start by discussing the applicability and success rate of the approach, the advantages it has over non-automated forecasts, and how mining our trained neural network provides a fresh look into the mechanisms behind magnetic energy storage and release.

  2. Convolutional neural networks with balanced batches for facial expressions recognition

    Science.gov (United States)

    Battini Sönmez, Elena; Cangelosi, Angelo

    2017-03-01

    This paper considers the issue of fully automatic emotion classification on 2D faces. In spite of the great effort done in recent years, traditional machine learning approaches based on hand-crafted feature extraction followed by the classification stage failed to develop a real-time automatic facial expression recognition system. The proposed architecture uses Convolutional Neural Networks (CNN), which are built as a collection of interconnected processing elements to simulate the brain of human beings. The basic idea of CNNs is to learn a hierarchical representation of the input data, which results in a better classification performance. In this work we present a block-based CNN algorithm, which uses noise, as data augmentation technique, and builds batches with a balanced number of samples per class. The proposed architecture is a very simple yet powerful CNN, which can yield state-of-the-art accuracy on the very competitive benchmark algorithm of the Extended Cohn Kanade database.

  3. Fully Convolutional Network Based Shadow Extraction from GF-2 Imagery

    Science.gov (United States)

    Li, Z.; Cai, G.; Ren, H.

    2018-04-01

    There are many shadows on the high spatial resolution satellite images, especially in the urban areas. Although shadows on imagery severely affect the information extraction of land cover or land use, they provide auxiliary information for building extraction which is hard to achieve a satisfactory accuracy through image classification itself. This paper focused on the method of building shadow extraction by designing a fully convolutional network and training samples collected from GF-2 satellite imagery in the urban region of Changchun city. By means of spatial filtering and calculation of adjacent relationship along the sunlight direction, the small patches from vegetation or bridges have been eliminated from the preliminary extracted shadows. Finally, the building shadows were separated. The extracted building shadow information from the proposed method in this paper was compared with the results from the traditional object-oriented supervised classification algorihtms. It showed that the deep learning network approach can improve the accuracy to a large extent.

  4. Fully convolutional network with cluster for semantic segmentation

    Science.gov (United States)

    Ma, Xiao; Chen, Zhongbi; Zhang, Jianlin

    2018-04-01

    At present, image semantic segmentation technology has been an active research topic for scientists in the field of computer vision and artificial intelligence. Especially, the extensive research of deep neural network in image recognition greatly promotes the development of semantic segmentation. This paper puts forward a method based on fully convolutional network, by cluster algorithm k-means. The cluster algorithm using the image's low-level features and initializing the cluster centers by the super-pixel segmentation is proposed to correct the set of points with low reliability, which are mistakenly classified in great probability, by the set of points with high reliability in each clustering regions. This method refines the segmentation of the target contour and improves the accuracy of the image segmentation.

  5. Real Time Eye Detector with Cascaded Convolutional Neural Networks

    Directory of Open Access Journals (Sweden)

    Bin Li

    2018-01-01

    Full Text Available An accurate and efficient eye detector is essential for many computer vision applications. In this paper, we present an efficient method to evaluate the eye location from facial images. First, a group of candidate regions with regional extreme points is quickly proposed; then, a set of convolution neural networks (CNNs is adopted to determine the most likely eye region and classify the region as left or right eye; finally, the center of the eye is located with other CNNs. In the experiments using GI4E, BioID, and our datasets, our method attained a detection accuracy which is comparable to existing state-of-the-art methods; meanwhile, our method was faster and adaptable to variations of the images, including external light changes, facial occlusion, and changes in image modality.

  6. Plane-wave decomposition by spherical-convolution microphone array

    Science.gov (United States)

    Rafaely, Boaz; Park, Munhum

    2004-05-01

    Reverberant sound fields are widely studied, as they have a significant influence on the acoustic performance of enclosures in a variety of applications. For example, the intelligibility of speech in lecture rooms, the quality of music in auditoria, the noise level in offices, and the production of 3D sound in living rooms are all affected by the enclosed sound field. These sound fields are typically studied through frequency response measurements or statistical measures such as reverberation time, which do not provide detailed spatial information. The aim of the work presented in this seminar is the detailed analysis of reverberant sound fields. A measurement and analysis system based on acoustic theory and signal processing, designed around a spherical microphone array, is presented. Detailed analysis is achieved by decomposition of the sound field into waves, using spherical Fourier transform and spherical convolution. The presentation will include theoretical review, simulation studies, and initial experimental results.

  7. Deep learning with convolutional neural network in radiology.

    Science.gov (United States)

    Yasaka, Koichiro; Akai, Hiroyuki; Kunimatsu, Akira; Kiryu, Shigeru; Abe, Osamu

    2018-04-01

    Deep learning with a convolutional neural network (CNN) is gaining attention recently for its high performance in image recognition. Images themselves can be utilized in a learning process with this technique, and feature extraction in advance of the learning process is not required. Important features can be automatically learned. Thanks to the development of hardware and software in addition to techniques regarding deep learning, application of this technique to radiological images for predicting clinically useful information, such as the detection and the evaluation of lesions, etc., are beginning to be investigated. This article illustrates basic technical knowledge regarding deep learning with CNNs along the actual course (collecting data, implementing CNNs, and training and testing phases). Pitfalls regarding this technique and how to manage them are also illustrated. We also described some advanced topics of deep learning, results of recent clinical studies, and the future directions of clinical application of deep learning techniques.

  8. DeepNAT: Deep convolutional neural network for segmenting neuroanatomy.

    Science.gov (United States)

    Wachinger, Christian; Reuter, Martin; Klein, Tassilo

    2018-04-15

    We introduce DeepNAT, a 3D Deep convolutional neural network for the automatic segmentation of NeuroAnaTomy in T1-weighted magnetic resonance images. DeepNAT is an end-to-end learning-based approach to brain segmentation that jointly learns an abstract feature representation and a multi-class classification. We propose a 3D patch-based approach, where we do not only predict the center voxel of the patch but also neighbors, which is formulated as multi-task learning. To address a class imbalance problem, we arrange two networks hierarchically, where the first one separates foreground from background, and the second one identifies 25 brain structures on the foreground. Since patches lack spatial context, we augment them with coordinates. To this end, we introduce a novel intrinsic parameterization of the brain volume, formed by eigenfunctions of the Laplace-Beltrami operator. As network architecture, we use three convolutional layers with pooling, batch normalization, and non-linearities, followed by fully connected layers with dropout. The final segmentation is inferred from the probabilistic output of the network with a 3D fully connected conditional random field, which ensures label agreement between close voxels. The roughly 2.7million parameters in the network are learned with stochastic gradient descent. Our results show that DeepNAT compares favorably to state-of-the-art methods. Finally, the purely learning-based method may have a high potential for the adaptation to young, old, or diseased brains by fine-tuning the pre-trained network with a small training sample on the target application, where the availability of larger datasets with manual annotations may boost the overall segmentation accuracy in the future. Copyright © 2017 Elsevier Inc. All rights reserved.

  9. FULLY CONVOLUTIONAL NETWORKS FOR GROUND CLASSIFICATION FROM LIDAR POINT CLOUDS

    Directory of Open Access Journals (Sweden)

    A. Rizaldy

    2018-05-01

    Full Text Available Deep Learning has been massively used for image classification in recent years. The use of deep learning for ground classification from LIDAR point clouds has also been recently studied. However, point clouds need to be converted into an image in order to use Convolutional Neural Networks (CNNs. In state-of-the-art techniques, this conversion is slow because each point is converted into a separate image. This approach leads to highly redundant computation during conversion and classification. The goal of this study is to design a more efficient data conversion and ground classification. This goal is achieved by first converting the whole point cloud into a single image. The classification is then performed by a Fully Convolutional Network (FCN, a modified version of CNN designed for pixel-wise image classification. The proposed method is significantly faster than state-of-the-art techniques. On the ISPRS Filter Test dataset, it is 78 times faster for conversion and 16 times faster for classification. Our experimental analysis on the same dataset shows that the proposed method results in 5.22 % of total error, 4.10 % of type I error, and 15.07 % of type II error. Compared to the previous CNN-based technique and LAStools software, the proposed method reduces the total error and type I error (while type II error is slightly higher. The method was also tested on a very high point density LIDAR point clouds resulting in 4.02 % of total error, 2.15 % of type I error and 6.14 % of type II error.

  10. Fully Convolutional Networks for Ground Classification from LIDAR Point Clouds

    Science.gov (United States)

    Rizaldy, A.; Persello, C.; Gevaert, C. M.; Oude Elberink, S. J.

    2018-05-01

    Deep Learning has been massively used for image classification in recent years. The use of deep learning for ground classification from LIDAR point clouds has also been recently studied. However, point clouds need to be converted into an image in order to use Convolutional Neural Networks (CNNs). In state-of-the-art techniques, this conversion is slow because each point is converted into a separate image. This approach leads to highly redundant computation during conversion and classification. The goal of this study is to design a more efficient data conversion and ground classification. This goal is achieved by first converting the whole point cloud into a single image. The classification is then performed by a Fully Convolutional Network (FCN), a modified version of CNN designed for pixel-wise image classification. The proposed method is significantly faster than state-of-the-art techniques. On the ISPRS Filter Test dataset, it is 78 times faster for conversion and 16 times faster for classification. Our experimental analysis on the same dataset shows that the proposed method results in 5.22 % of total error, 4.10 % of type I error, and 15.07 % of type II error. Compared to the previous CNN-based technique and LAStools software, the proposed method reduces the total error and type I error (while type II error is slightly higher). The method was also tested on a very high point density LIDAR point clouds resulting in 4.02 % of total error, 2.15 % of type I error and 6.14 % of type II error.

  11. A frequency bin-wise nonlinear masking algorithm in convolutive mixtures for speech segregation.

    Science.gov (United States)

    Chi, Tai-Shih; Huang, Ching-Wen; Chou, Wen-Sheng

    2012-05-01

    A frequency bin-wise nonlinear masking algorithm is proposed in the spectrogram domain for speech segregation in convolutive mixtures. The contributive weight from each speech source to a time-frequency unit of the mixture spectrogram is estimated by a nonlinear function based on location cues. For each sound source, a non-binary mask is formed from the estimated weights and is multiplied to the mixture spectrogram to extract the sound. Head-related transfer functions (HRTFs) are used to simulate convolutive sound mixtures perceived by listeners. Simulation results show our proposed method outperforms convolutive independent component analysis and degenerate unmixing and estimation technique methods in almost all test conditions.

  12. Development and application of deep convolutional neural network in target detection

    Science.gov (United States)

    Jiang, Xiaowei; Wang, Chunping; Fu, Qiang

    2018-04-01

    With the development of big data and algorithms, deep convolution neural networks with more hidden layers have more powerful feature learning and feature expression ability than traditional machine learning methods, making artificial intelligence surpass human level in many fields. This paper first reviews the development and application of deep convolutional neural networks in the field of object detection in recent years, then briefly summarizes and ponders some existing problems in the current research, and the future development of deep convolutional neural network is prospected.

  13. MR-based synthetic CT generation using a deep convolutional neural network method.

    Science.gov (United States)

    Han, Xiao

    2017-04-01

    Interests have been rapidly growing in the field of radiotherapy to replace CT with magnetic resonance imaging (MRI), due to superior soft tissue contrast offered by MRI and the desire to reduce unnecessary radiation dose. MR-only radiotherapy also simplifies clinical workflow and avoids uncertainties in aligning MR with CT. Methods, however, are needed to derive CT-equivalent representations, often known as synthetic CT (sCT), from patient MR images for dose calculation and DRR-based patient positioning. Synthetic CT estimation is also important for PET attenuation correction in hybrid PET-MR systems. We propose in this work a novel deep convolutional neural network (DCNN) method for sCT generation and evaluate its performance on a set of brain tumor patient images. The proposed method builds upon recent developments of deep learning and convolutional neural networks in the computer vision literature. The proposed DCNN model has 27 convolutional layers interleaved with pooling and unpooling layers and 35 million free parameters, which can be trained to learn a direct end-to-end mapping from MR images to their corresponding CTs. Training such a large model on our limited data is made possible through the principle of transfer learning and by initializing model weights from a pretrained model. Eighteen brain tumor patients with both CT and T1-weighted MR images are used as experimental data and a sixfold cross-validation study is performed. Each sCT generated is compared against the real CT image of the same patient on a voxel-by-voxel basis. Comparison is also made with respect to an atlas-based approach that involves deformable atlas registration and patch-based atlas fusion. The proposed DCNN method produced a mean absolute error (MAE) below 85 HU for 13 of the 18 test subjects. The overall average MAE was 84.8 ± 17.3 HU for all subjects, which was found to be significantly better than the average MAE of 94.5 ± 17.8 HU for the atlas-based method. The DCNN

  14. Sustainability Analysis and Market Demand Estimation in the Retail Industry through a Convolutional Neural Network

    Directory of Open Access Journals (Sweden)

    Luyao Wang

    2018-05-01

    Full Text Available The Chinese retail industry is expected to grow dramatically over the next few years, owing to the rapid increase in purchasing power of Chinese consumers. Retail managers should analyze the market demands and avoid dull sales to promote the sustainable development of the retail industry. Economic sustainability in the retail industry, which refers to a suitable return of investment, requires the implementation of precise product allocation strategies in different regions. This study proposed a hybrid model to evaluate economic sustainability in the preparation of goods of retail shops on the basis of market demand evaluation. Through a grid-based convolutional neural network, a regression model was first established to model the relationship between consumer distribution and the potential market demand. Then, another model was proposed to evaluate the sustainability among regions based on their supply-demand analysis. An experiment was conducted based on the actual sales data of retail shops in Guiyang, China. Results showed an immense diversity of sustainability in the entire city and three classes of regions were distinguished, namely, high, moderate, and limited. Our model was proven to be effective in the sustainability evaluation of supply and demand in the retail industry after validation showed that its accuracy reached 92.8%.

  15. Compression of deep convolutional neural network for computer-aided diagnosis of masses in digital breast tomosynthesis

    Science.gov (United States)

    Samala, Ravi K.; Chan, Heang-Ping; Hadjiiski, Lubomir; Helvie, Mark A.; Richter, Caleb; Cha, Kenny

    2018-02-01

    Deep-learning models are highly parameterized, causing difficulty in inference and transfer learning. We propose a layered pathway evolution method to compress a deep convolutional neural network (DCNN) for classification of masses in DBT while maintaining the classification accuracy. Two-stage transfer learning was used to adapt the ImageNet-trained DCNN to mammography and then to DBT. In the first-stage transfer learning, transfer learning from ImageNet trained DCNN was performed using mammography data. In the second-stage transfer learning, the mammography-trained DCNN was trained on the DBT data using feature extraction from fully connected layer, recursive feature elimination and random forest classification. The layered pathway evolution encapsulates the feature extraction to the classification stages to compress the DCNN. Genetic algorithm was used in an iterative approach with tournament selection driven by count-preserving crossover and mutation to identify the necessary nodes in each convolution layer while eliminating the redundant nodes. The DCNN was reduced by 99% in the number of parameters and 95% in mathematical operations in the convolutional layers. The lesion-based area under the receiver operating characteristic curve on an independent DBT test set from the original and the compressed network resulted in 0.88+/-0.05 and 0.90+/-0.04, respectively. The difference did not reach statistical significance. We demonstrated a DCNN compression approach without additional fine-tuning or loss of performance for classification of masses in DBT. The approach can be extended to other DCNNs and transfer learning tasks. An ensemble of these smaller and focused DCNNs has the potential to be used in multi-target transfer learning.

  16. Soil Moisture Retrieval Using Convolutional Neural Networks: Application to Passive Microwave Remote Sensing

    Science.gov (United States)

    Hu, Z.; Xu, L.; Yu, B.

    2018-04-01

    A empirical model is established to analyse the daily retrieval of soil moisture from passive microwave remote sensing using convolutional neural networks (CNN). Soil moisture plays an important role in the water cycle. However, with the rapidly increasing of the acquiring technology for remotely sensed data, it's a hard task for remote sensing practitioners to find a fast and convenient model to deal with the massive data. In this paper, the AMSR-E brightness temperatures are used to train CNN for the prediction of the European centre for medium-range weather forecasts (ECMWF) model. Compared with the classical inversion methods, the deep learning-based method is more suitable for global soil moisture retrieval. It is very well supported by graphics processing unit (GPU) acceleration, which can meet the demand of massive data inversion. Once the model trained, a global soil moisture map can be predicted in less than 10 seconds. What's more, the method of soil moisture retrieval based on deep learning can learn the complex texture features from the big remote sensing data. In this experiment, the results demonstrates that the CNN deployed to retrieve global soil moisture can achieve a better performance than the support vector regression (SVR) for soil moisture retrieval.

  17. SOIL MOISTURE RETRIEVAL USING CONVOLUTIONAL NEURAL NETWORKS: APPLICATION TO PASSIVE MICROWAVE REMOTE SENSING

    Directory of Open Access Journals (Sweden)

    Z. Hu

    2018-04-01

    Full Text Available A empirical model is established to analyse the daily retrieval of soil moisture from passive microwave remote sensing using convolutional neural networks (CNN. Soil moisture plays an important role in the water cycle. However, with the rapidly increasing of the acquiring technology for remotely sensed data, it's a hard task for remote sensing practitioners to find a fast and convenient model to deal with the massive data. In this paper, the AMSR-E brightness temperatures are used to train CNN for the prediction of the European centre for medium-range weather forecasts (ECMWF model. Compared with the classical inversion methods, the deep learning-based method is more suitable for global soil moisture retrieval. It is very well supported by graphics processing unit (GPU acceleration, which can meet the demand of massive data inversion. Once the model trained, a global soil moisture map can be predicted in less than 10 seconds. What's more, the method of soil moisture retrieval based on deep learning can learn the complex texture features from the big remote sensing data. In this experiment, the results demonstrates that the CNN deployed to retrieve global soil moisture can achieve a better performance than the support vector regression (SVR for soil moisture retrieval.

  18. Subsidence feature discrimination using deep convolutional neral networks in synthetic aperture radar imagery

    CSIR Research Space (South Africa)

    Schwegmann, Colin P

    2017-07-01

    Full Text Available International Geoscience and Remote Sensing Symposium (IGARSS), 23-28 July 2017, Fort Worth, TX, USA SUBSIDENCE FEATURE DISCRIMINATION USING DEEP CONVOLUTIONAL NEURAL NETWORKS IN SYNTHETIC APERTURE RADAR IMAGERY Schwegmann, Colin P Kleynhans, Waldo...

  19. Single Image Super-Resolution Based on Multi-Scale Competitive Convolutional Neural Network.

    Science.gov (United States)

    Du, Xiaofeng; Qu, Xiaobo; He, Yifan; Guo, Di

    2018-03-06

    Deep convolutional neural networks (CNNs) are successful in single-image super-resolution. Traditional CNNs are limited to exploit multi-scale contextual information for image reconstruction due to the fixed convolutional kernel in their building modules. To restore various scales of image details, we enhance the multi-scale inference capability of CNNs by introducing competition among multi-scale convolutional filters, and build up a shallow network under limited computational resources. The proposed network has the following two advantages: (1) the multi-scale convolutional kernel provides the multi-context for image super-resolution, and (2) the maximum competitive strategy adaptively chooses the optimal scale of information for image reconstruction. Our experimental results on image super-resolution show that the performance of the proposed network outperforms the state-of-the-art methods.

  20. A Revised Piecewise Linear Recursive Convolution FDTD Method for Magnetized Plasmas

    International Nuclear Information System (INIS)

    Liu Song; Zhong Shuangying; Liu Shaobin

    2005-01-01

    The piecewise linear recursive convolution (PLRC) finite-different time-domain (FDTD) method improves accuracy over the original recursive convolution (RC) FDTD approach and current density convolution (JEC) but retains their advantages in speed and efficiency. This paper describes a revised piecewise linear recursive convolution PLRC-FDTD formulation for magnetized plasma which incorporates both anisotropy and frequency dispersion at the same time, enabling the transient analysis of magnetized plasma media. The technique is illustrated by numerical simulations of the reflection and transmission coefficients through a magnetized plasma layer. The results show that the revised PLRC-FDTD method has improved the accuracy over the original RC FDTD method and JEC FDTD method

  1. Deep Fully Convolutional Networks for the Detection of Informal Settlements in VHR Images

    NARCIS (Netherlands)

    Persello, Claudio; Stein, Alfred

    2017-01-01

    This letter investigates fully convolutional networks (FCNs) for the detection of informal settlements in very high resolution (VHR) satellite images. Informal settlements or slums are proliferating in developing countries and their detection and classification provides vital information for

  2. Fourier transform and mean quadratic variation of Bernoulli convolution on homogeneous Cantor set

    Energy Technology Data Exchange (ETDEWEB)

    Yu Zuguo E-mail: yuzg@hotmail.comz.yu

    2004-07-01

    For the Bernoulli convolution on homogeneous Cantor set, under some condition, it is proved that the mean quadratic variation and the average of Fourier transform of this measure are bounded above and below.

  3. Directional Radiometry and Radiative Transfer: the Convoluted Path From Centuries-old Phenomenology to Physical Optics

    Science.gov (United States)

    Mishchenko, Michael I.

    2014-01-01

    This Essay traces the centuries-long history of the phenomenological disciplines of directional radiometry and radiative transfer in turbid media, discusses their fundamental weaknesses, and outlines the convoluted process of their conversion into legitimate branches of physical optics.

  4. The neuro vector engine : flexibility to improve convolutional net efficiency for wearable vision

    NARCIS (Netherlands)

    Peemen, M.C.J.; Shi, R.; Lal, S.; Juurlink, B.H.H.; Mesman, B.; Corporaal, H.

    2016-01-01

    Deep Convolutional Networks (ConvNets) are currently superior in benchmark performance, but the associated demands on computation and data transfer prohibit straightforward mapping on energy constrained wearable platforms. The computational burden can be overcome by dedicated hardware accelerators,

  5. Method for assessing the probability of accumulated doses from an intermittent source using the convolution technique

    International Nuclear Information System (INIS)

    Coleman, J.H.

    1980-10-01

    A technique is discussed for computing the probability distribution of the accumulated dose received by an arbitrary receptor resulting from several single releases from an intermittent source. The probability density of the accumulated dose is the convolution of the probability densities of doses from the intermittent releases. Emissions are not assumed to be constant over the brief release period. The fast fourier transform is used in the calculation of the convolution

  6. Deep Galaxy: Classification of Galaxies based on Deep Convolutional Neural Networks

    OpenAIRE

    Khalifa, Nour Eldeen M.; Taha, Mohamed Hamed N.; Hassanien, Aboul Ella; Selim, I. M.

    2017-01-01

    In this paper, a deep convolutional neural network architecture for galaxies classification is presented. The galaxy can be classified based on its features into main three categories Elliptical, Spiral, and Irregular. The proposed deep galaxies architecture consists of 8 layers, one main convolutional layer for features extraction with 96 filters, followed by two principles fully connected layers for classification. It is trained over 1356 images and achieved 97.272% in testing accuracy. A c...

  7. Is Kinesio Taping to Generate Skin Convolutions Effective for Increasing Local Blood Circulation?

    OpenAIRE

    Yang, Jae-Man; Lee, Jung-Hoon

    2018-01-01

    Background It is unclear whether traditional application of Kinesio taping, which produces wrinkles in the skin, is effective for improving blood circulation. This study investigated local skin temperature changes after the application of an elastic therapeutic tape using convolution and non-convolution taping methods (CTM/NCTM). Material/Methods Twenty-eight pain-free men underwent CTM and NCTM randomly applied to the right and left sides of the lower back. Using infrared thermography, skin ...

  8. Isointense infant brain MRI segmentation with a dilated convolutional neural network

    OpenAIRE

    Moeskops, Pim; Pluim, Josien P. W.

    2017-01-01

    Quantitative analysis of brain MRI at the age of 6 months is difficult because of the limited contrast between white matter and gray matter. In this study, we use a dilated triplanar convolutional neural network in combination with a non-dilated 3D convolutional neural network for the segmentation of white matter, gray matter and cerebrospinal fluid in infant brain MR images, as provided by the MICCAI grand challenge on 6-month infant brain MRI segmentation.

  9. Experimental study of current loss and plasma formation in the Z machine post-hole convolute

    Directory of Open Access Journals (Sweden)

    M. R. Gomez

    2017-01-01

    Full Text Available The Z pulsed-power generator at Sandia National Laboratories drives high energy density physics experiments with load currents of up to 26 MA. Z utilizes a double post-hole convolute to combine the current from four parallel magnetically insulated transmission lines into a single transmission line just upstream of the load. Current loss is observed in most experiments and is traditionally attributed to inefficient convolute performance. The apparent loss current varies substantially for z-pinch loads with different inductance histories; however, a similar convolute impedance history is observed for all load types. This paper details direct spectroscopic measurements of plasma density, temperature, and apparent and actual plasma closure velocities within the convolute. Spectral measurements indicate a correlation between impedance collapse and plasma formation in the convolute. Absorption features in the spectra show the convolute plasma consists primarily of hydrogen, which likely forms from desorbed electrode contaminant species such as H_{2}O, H_{2}, and hydrocarbons. Plasma densities increase from 1×10^{16}  cm^{−3} (level of detectability just before peak current to over 1×10^{17}  cm^{−3} at stagnation (tens of ns later. The density seems to be highest near the cathode surface, with an apparent cathode to anode plasma velocity in the range of 35–50  cm/μs. Similar plasma conditions and convolute impedance histories are observed in experiments with high and low losses, suggesting that losses are driven largely by load dynamics, which determine the voltage on the convolute.

  10. The convolution integral for the forward-backward asymmetry in e+e- annihilation

    International Nuclear Information System (INIS)

    Bardin, D.; Bilenky, M.; Chizhov, A.; Sazonov, A.; Sedykh, Yu.; Riemann, T.; Sachwitz, M.

    1989-01-01

    The complete convolution integral for the forward-backward asymmetry in A FB in e + e - annihilation is obtained in order O(α) with soft photon exponentiation. The influence of these QED corrections on A FB in the vicinity of the Z peak is discussed. The results are used to comment on a recent ad hoc ansatz using convolution weights derived for the total cross section. (orig.)

  11. Esophagus segmentation in CT via 3D fully convolutional neural network and random walk.

    Science.gov (United States)

    Fechter, Tobias; Adebahr, Sonja; Baltas, Dimos; Ben Ayed, Ismail; Desrosiers, Christian; Dolz, Jose

    2017-12-01

    Precise delineation of organs at risk is a crucial task in radiotherapy treatment planning for delivering high doses to the tumor while sparing healthy tissues. In recent years, automated segmentation methods have shown an increasingly high performance for the delineation of various anatomical structures. However, this task remains challenging for organs like the esophagus, which have a versatile shape and poor contrast to neighboring tissues. For human experts, segmenting the esophagus from CT images is a time-consuming and error-prone process. To tackle these issues, we propose a random walker approach driven by a 3D fully convolutional neural network (CNN) to automatically segment the esophagus from CT images. First, a soft probability map is generated by the CNN. Then, an active contour model (ACM) is fitted to the CNN soft probability map to get a first estimation of the esophagus location. The outputs of the CNN and ACM are then used in conjunction with a probability model based on CT Hounsfield (HU) values to drive the random walker. Training and evaluation were done on 50 CTs from two different datasets, with clinically used peer-reviewed esophagus contours. Results were assessed regarding spatial overlap and shape similarity. The esophagus contours generated by the proposed algorithm showed a mean Dice coefficient of 0.76 ± 0.11, an average symmetric square distance of 1.36 ± 0.90 mm, and an average Hausdorff distance of 11.68 ± 6.80, compared to the reference contours. These results translate to a very good agreement with reference contours and an increase in accuracy compared to existing methods. Furthermore, when considering the results reported in the literature for the publicly available Synapse dataset, our method outperformed all existing approaches, which suggests that the proposed method represents the current state-of-the-art for automatic esophagus segmentation. We show that a CNN can yield accurate estimations of esophagus location, and that

  12. Minimal-memory realization of pearl-necklace encoders of general quantum convolutional codes

    International Nuclear Information System (INIS)

    Houshmand, Monireh; Hosseini-Khayat, Saied

    2011-01-01

    Quantum convolutional codes, like their classical counterparts, promise to offer higher error correction performance than block codes of equivalent encoding complexity, and are expected to find important applications in reliable quantum communication where a continuous stream of qubits is transmitted. Grassl and Roetteler devised an algorithm to encode a quantum convolutional code with a ''pearl-necklace'' encoder. Despite their algorithm's theoretical significance as a neat way of representing quantum convolutional codes, it is not well suited to practical realization. In fact, there is no straightforward way to implement any given pearl-necklace structure. This paper closes the gap between theoretical representation and practical implementation. In our previous work, we presented an efficient algorithm to find a minimal-memory realization of a pearl-necklace encoder for Calderbank-Shor-Steane (CSS) convolutional codes. This work is an extension of our previous work and presents an algorithm for turning a pearl-necklace encoder for a general (non-CSS) quantum convolutional code into a realizable quantum convolutional encoder. We show that a minimal-memory realization depends on the commutativity relations between the gate strings in the pearl-necklace encoder. We find a realization by means of a weighted graph which details the noncommutative paths through the pearl necklace. The weight of the longest path in this graph is equal to the minimal amount of memory needed to implement the encoder. The algorithm has a polynomial-time complexity in the number of gate strings in the pearl-necklace encoder.

  13. On the Reduction of Computational Complexity of Deep Convolutional Neural Networks

    Directory of Open Access Journals (Sweden)

    Partha Maji

    2018-04-01

    Full Text Available Deep convolutional neural networks (ConvNets, which are at the heart of many new emerging applications, achieve remarkable performance in audio and visual recognition tasks. Unfortunately, achieving accuracy often implies significant computational costs, limiting deployability. In modern ConvNets it is typical for the convolution layers to consume the vast majority of computational resources during inference. This has made the acceleration of these layers an important research area in academia and industry. In this paper, we examine the effects of co-optimizing the internal structures of the convolutional layers and underlying implementation of fundamental convolution operation. We demonstrate that a combination of these methods can have a big impact on the overall speedup of a ConvNet, achieving a ten-fold increase over baseline. We also introduce a new class of fast one-dimensional (1D convolutions for ConvNets using the Toom–Cook algorithm. We show that our proposed scheme is mathematically well-grounded, robust, and does not require any time-consuming retraining, while still achieving speedups solely from convolutional layers with no loss in baseline accuracy.

  14. A convolution method for predicting mean treatment dose including organ motion at imaging

    International Nuclear Information System (INIS)

    Booth, J.T.; Zavgorodni, S.F.; Royal Adelaide Hospital, SA

    2000-01-01

    Full text: The random treatment delivery errors (organ motion and set-up error) can be incorporated into the treatment planning software using a convolution method. Mean treatment dose is computed as the convolution of a static dose distribution with a variation kernel. Typically this variation kernel is Gaussian with variance equal to the sum of the organ motion and set-up error variances. We propose a novel variation kernel for the convolution technique that additionally considers the position of the mobile organ in the planning CT image. The systematic error of organ position in the planning CT image can be considered random for each patient over a population. Thus the variance of the variation kernel will equal the sum of treatment delivery variance and organ motion variance at planning for the population of treatments. The kernel is extended to deal with multiple pre-treatment CT scans to improve tumour localisation for planning. Mean treatment doses calculated with the convolution technique are compared to benchmark Monte Carlo (MC) computations. Calculations of mean treatment dose using the convolution technique agreed with MC results for all cases to better than ± 1 Gy in the planning treatment volume for a prescribed 60 Gy treatment. Convolution provides a quick method of incorporating random organ motion (captured in the planning CT image and during treatment delivery) and random set-up errors directly into the dose distribution. Copyright (2000) Australasian College of Physical Scientists and Engineers in Medicine

  15. Performance evaluation of image denoising developed using convolutional denoising autoencoders in chest radiography

    Science.gov (United States)

    Lee, Donghoon; Choi, Sunghoon; Kim, Hee-Joung

    2018-03-01

    When processing medical images, image denoising is an important pre-processing step. Various image denoising algorithms have been developed in the past few decades. Recently, image denoising using the deep learning method has shown excellent performance compared to conventional image denoising algorithms. In this study, we introduce an image denoising technique based on a convolutional denoising autoencoder (CDAE) and evaluate clinical applications by comparing existing image denoising algorithms. We train the proposed CDAE model using 3000 chest radiograms training data. To evaluate the performance of the developed CDAE model, we compare it with conventional denoising algorithms including median filter, total variation (TV) minimization, and non-local mean (NLM) algorithms. Furthermore, to verify the clinical effectiveness of the developed denoising model with CDAE, we investigate the performance of the developed denoising algorithm on chest radiograms acquired from real patients. The results demonstrate that the proposed denoising algorithm developed using CDAE achieves a superior noise-reduction effect in chest radiograms compared to TV minimization and NLM algorithms, which are state-of-the-art algorithms for image noise reduction. For example, the peak signal-to-noise ratio and structure similarity index measure of CDAE were at least 10% higher compared to conventional denoising algorithms. In conclusion, the image denoising algorithm developed using CDAE effectively eliminated noise without loss of information on anatomical structures in chest radiograms. It is expected that the proposed denoising algorithm developed using CDAE will be effective for medical images with microscopic anatomical structures, such as terminal bronchioles.

  16. Deep-HiTS: Rotation Invariant Convolutional Neural Network for Transient Detection

    Science.gov (United States)

    Cabrera-Vives, Guillermo; Reyes, Ignacio; Förster, Francisco; Estévez, Pablo A.; Maureira, Juan-Carlos

    2017-02-01

    We introduce Deep-HiTS, a rotation-invariant convolutional neural network (CNN) model for classifying images of transient candidates into artifacts or real sources for the High cadence Transient Survey (HiTS). CNNs have the advantage of learning the features automatically from the data while achieving high performance. We compare our CNN model against a feature engineering approach using random forests (RFs). We show that our CNN significantly outperforms the RF model, reducing the error by almost half. Furthermore, for a fixed number of approximately 2000 allowed false transient candidates per night, we are able to reduce the misclassified real transients by approximately one-fifth. To the best of our knowledge, this is the first time CNNs have been used to detect astronomical transient events. Our approach will be very useful when processing images from next generation instruments such as the Large Synoptic Survey Telescope. We have made all our code and data available to the community for the sake of allowing further developments and comparisons at https://github.com/guille-c/Deep-HiTS. Deep-HiTS is licensed under the terms of the GNU General Public License v3.0.

  17. The Use of Convolutional Neural Network in Relating Precipitation to Circulation

    Science.gov (United States)

    Pan, B.; Hsu, K. L.; AghaKouchak, A.; Sorooshian, S.

    2017-12-01

    Precipitation prediction in dynamical weather and climate models depends on 1) the predictability of pressure or geopotential height for the forecasting period and 2) the successive work of interpreting the pressure field in terms of precipitation events. The later task is represented as parameterization schemes in numerical models, where detailed computing inevitably blurs the hidden cause-and-effect relationship in precipitation generation. The "big data" provided by numerical simulation, reanalysis and observation networks requires better causation analysis for people to digest and realize their use. While classic synoptical analysis methods are very-often insufficient for spatially distributed high dimensional data, a Convolutional Neural Network(CNN) is developed here to directly relate precipitation with circulation. Case study carried over west coast United States during boreal winter showed that CNN can locate and capture key pressure zones of different structures to project precipitation spatial distribution with high accuracy across hourly to monthly scales. This direct connection between atmospheric circulation and precipitation offers a probe for attributing precipitation to the coverage, location, intensity and spatial structure of characteristic pressure zones, which can be used for model diagnosis and improvement.

  18. Epithelium-Stroma Classification via Convolutional Neural Networks and Unsupervised Domain Adaptation in Histopathological Images.

    Science.gov (United States)

    Huang, Yue; Zheng, Han; Liu, Chi; Ding, Xinghao; Rohde, Gustavo K

    2017-11-01

    Epithelium-stroma classification is a necessary preprocessing step in histopathological image analysis. Current deep learning based recognition methods for histology data require collection of large volumes of labeled data in order to train a new neural network when there are changes to the image acquisition procedure. However, it is extremely expensive for pathologists to manually label sufficient volumes of data for each pathology study in a professional manner, which results in limitations in real-world applications. A very simple but effective deep learning method, that introduces the concept of unsupervised domain adaptation to a simple convolutional neural network (CNN), has been proposed in this paper. Inspired by transfer learning, our paper assumes that the training data and testing data follow different distributions, and there is an adaptation operation to more accurately estimate the kernels in CNN in feature extraction, in order to enhance performance by transferring knowledge from labeled data in source domain to unlabeled data in target domain. The model has been evaluated using three independent public epithelium-stroma datasets by cross-dataset validations. The experimental results demonstrate that for epithelium-stroma classification, the proposed framework outperforms the state-of-the-art deep neural network model, and it also achieves better performance than other existing deep domain adaptation methods. The proposed model can be considered to be a better option for real-world applications in histopathological image analysis, since there is no longer a requirement for large-scale labeled data in each specified domain.

  19. Deep Learning at Chest Radiography: Automated Classification of Pulmonary Tuberculosis by Using Convolutional Neural Networks.

    Science.gov (United States)

    Lakhani, Paras; Sundaram, Baskaran

    2017-08-01

    Purpose To evaluate the efficacy of deep convolutional neural networks (DCNNs) for detecting tuberculosis (TB) on chest radiographs. Materials and Methods Four deidentified HIPAA-compliant datasets were used in this study that were exempted from review by the institutional review board, which consisted of 1007 posteroanterior chest radiographs. The datasets were split into training (68.0%), validation (17.1%), and test (14.9%). Two different DCNNs, AlexNet and GoogLeNet, were used to classify the images as having manifestations of pulmonary TB or as healthy. Both untrained and pretrained networks on ImageNet were used, and augmentation with multiple preprocessing techniques. Ensembles were performed on the best-performing algorithms. For cases where the classifiers were in disagreement, an independent board-certified cardiothoracic radiologist blindly interpreted the images to evaluate a potential radiologist-augmented workflow. Receiver operating characteristic curves and areas under the curve (AUCs) were used to assess model performance by using the DeLong method for statistical comparison of receiver operating characteristic curves. Results The best-performing classifier had an AUC of 0.99, which was an ensemble of the AlexNet and GoogLeNet DCNNs. The AUCs of the pretrained models were greater than that of the untrained models (P chest radiography with an AUC of 0.99. A radiologist-augmented approach for cases where there was disagreement among the classifiers further improved accuracy. © RSNA, 2017.

  20. Evolutionary Design of Convolutional Neural Networks for Human Activity Recognition in Sensor-Rich Environments

    Science.gov (United States)

    2018-01-01

    Human activity recognition is a challenging problem for context-aware systems and applications. It is gaining interest due to the ubiquity of different sensor sources, wearable smart objects, ambient sensors, etc. This task is usually approached as a supervised machine learning problem, where a label is to be predicted given some input data, such as the signals retrieved from different sensors. For tackling the human activity recognition problem in sensor network environments, in this paper we propose the use of deep learning (convolutional neural networks) to perform activity recognition using the publicly available OPPORTUNITY dataset. Instead of manually choosing a suitable topology, we will let an evolutionary algorithm design the optimal topology in order to maximize the classification F1 score. After that, we will also explore the performance of committees of the models resulting from the evolutionary process. Results analysis indicates that the proposed model was able to perform activity recognition within a heterogeneous sensor network environment, achieving very high accuracies when tested with new sensor data. Based on all conducted experiments, the proposed neuroevolutionary system has proved to be able to systematically find a classification model which is capable of outperforming previous results reported in the state-of-the-art, showing that this approach is useful and improves upon previously manually-designed architectures. PMID:29690587

  1. Evolutionary Design of Convolutional Neural Networks for Human Activity Recognition in Sensor-Rich Environments

    Directory of Open Access Journals (Sweden)

    Alejandro Baldominos

    2018-04-01

    Full Text Available Human activity recognition is a challenging problem for context-aware systems and applications. It is gaining interest due to the ubiquity of different sensor sources, wearable smart objects, ambient sensors, etc. This task is usually approached as a supervised machine learning problem, where a label is to be predicted given some input data, such as the signals retrieved from different sensors. For tackling the human activity recognition problem in sensor network environments, in this paper we propose the use of deep learning (convolutional neural networks to perform activity recognition using the publicly available OPPORTUNITY dataset. Instead of manually choosing a suitable topology, we will let an evolutionary algorithm design the optimal topology in order to maximize the classification F1 score. After that, we will also explore the performance of committees of the models resulting from the evolutionary process. Results analysis indicates that the proposed model was able to perform activity recognition within a heterogeneous sensor network environment, achieving very high accuracies when tested with new sensor data. Based on all conducted experiments, the proposed neuroevolutionary system has proved to be able to systematically find a classification model which is capable of outperforming previous results reported in the state-of-the-art, showing that this approach is useful and improves upon previously manually-designed architectures.

  2. Evolutionary Design of Convolutional Neural Networks for Human Activity Recognition in Sensor-Rich Environments.

    Science.gov (United States)

    Baldominos, Alejandro; Saez, Yago; Isasi, Pedro

    2018-04-23

    Human activity recognition is a challenging problem for context-aware systems and applications. It is gaining interest due to the ubiquity of different sensor sources, wearable smart objects, ambient sensors, etc. This task is usually approached as a supervised machine learning problem, where a label is to be predicted given some input data, such as the signals retrieved from different sensors. For tackling the human activity recognition problem in sensor network environments, in this paper we propose the use of deep learning (convolutional neural networks) to perform activity recognition using the publicly available OPPORTUNITY dataset. Instead of manually choosing a suitable topology, we will let an evolutionary algorithm design the optimal topology in order to maximize the classification F1 score. After that, we will also explore the performance of committees of the models resulting from the evolutionary process. Results analysis indicates that the proposed model was able to perform activity recognition within a heterogeneous sensor network environment, achieving very high accuracies when tested with new sensor data. Based on all conducted experiments, the proposed neuroevolutionary system has proved to be able to systematically find a classification model which is capable of outperforming previous results reported in the state-of-the-art, showing that this approach is useful and improves upon previously manually-designed architectures.

  3. Image Classification Using Biomimetic Pattern Recognition with Convolutional Neural Networks Features

    Science.gov (United States)

    Huo, Guanying

    2017-01-01

    As a typical deep-learning model, Convolutional Neural Networks (CNNs) can be exploited to automatically extract features from images using the hierarchical structure inspired by mammalian visual system. For image classification tasks, traditional CNN models employ the softmax function for classification. However, owing to the limited capacity of the softmax function, there are some shortcomings of traditional CNN models in image classification. To deal with this problem, a new method combining Biomimetic Pattern Recognition (BPR) with CNNs is proposed for image classification. BPR performs class recognition by a union of geometrical cover sets in a high-dimensional feature space and therefore can overcome some disadvantages of traditional pattern recognition. The proposed method is evaluated on three famous image classification benchmarks, that is, MNIST, AR, and CIFAR-10. The classification accuracies of the proposed method for the three datasets are 99.01%, 98.40%, and 87.11%, respectively, which are much higher in comparison with the other four methods in most cases. PMID:28316614

  4. Three-dimensional fingerprint recognition by using convolution neural network

    Science.gov (United States)

    Tian, Qianyu; Gao, Nan; Zhang, Zonghua

    2018-01-01

    With the development of science and technology and the improvement of social information, fingerprint recognition technology has become a hot research direction and been widely applied in many actual fields because of its feasibility and reliability. The traditional two-dimensional (2D) fingerprint recognition method relies on matching feature points. This method is not only time-consuming, but also lost three-dimensional (3D) information of fingerprint, with the fingerprint rotation, scaling, damage and other issues, a serious decline in robustness. To solve these problems, 3D fingerprint has been used to recognize human being. Because it is a new research field, there are still lots of challenging problems in 3D fingerprint recognition. This paper presents a new 3D fingerprint recognition method by using a convolution neural network (CNN). By combining 2D fingerprint and fingerprint depth map into CNN, and then through another CNN feature fusion, the characteristics of the fusion complete 3D fingerprint recognition after classification. This method not only can preserve 3D information of fingerprints, but also solves the problem of CNN input. Moreover, the recognition process is simpler than traditional feature point matching algorithm. 3D fingerprint recognition rate by using CNN is compared with other fingerprint recognition algorithms. The experimental results show that the proposed 3D fingerprint recognition method has good recognition rate and robustness.

  5. Generalization error analysis: deep convolutional neural network in mammography

    Science.gov (United States)

    Richter, Caleb D.; Samala, Ravi K.; Chan, Heang-Ping; Hadjiiski, Lubomir; Cha, Kenny

    2018-02-01

    We conducted a study to gain understanding of the generalizability of deep convolutional neural networks (DCNNs) given their inherent capability to memorize data. We examined empirically a specific DCNN trained for classification of masses on mammograms. Using a data set of 2,454 lesions from 2,242 mammographic views, a DCNN was trained to classify masses into malignant and benign classes using transfer learning from ImageNet LSVRC-2010. We performed experiments with varying amounts of label corruption and types of pixel randomization to analyze the generalization error for the DCNN. Performance was evaluated using the area under the receiver operating characteristic curve (AUC) with an N-fold cross validation. Comparisons were made between the convergence times, the inference AUCs for both the training set and the test set of the original image patches without corruption, and the root-mean-squared difference (RMSD) in the layer weights of the DCNN trained with different amounts and methods of corruption. Our experiments observed trends which revealed that the DCNN overfitted by memorizing corrupted data. More importantly, this study improved our understanding of DCNN weight updates when learning new patterns or new labels. Although we used a specific classification task with the ImageNet as example, similar methods may be useful for analysis of the DCNN learning processes, especially those that employ transfer learning for medical image analysis where sample size is limited and overfitting risk is high.

  6. A discrete convolution kernel for No-DC MRI

    International Nuclear Information System (INIS)

    Zeng, Gengsheng L; Li, Ya

    2015-01-01

    An analytical inversion formula for the exponential Radon transform with an imaginary attenuation coefficient was developed in 2007 (2007 Inverse Problems 23 1963–71). The inversion formula in that paper suggested that it is possible to obtain an exact MRI (magnetic resonance imaging) image without acquiring low-frequency data. However, this un-measured low-frequency region (ULFR) in the k-space (which is the two-dimensional Fourier transform space in MRI terminology) must be very small. This current paper derives a FBP (filtered backprojection) algorithm based on You’s formula by suggesting a practical discrete convolution kernel. A point spread function is derived for this FBP algorithm. It is demonstrated that the derived FBP algorithm can have a larger ULFR than that in the 2007 paper. The significance of this paper is that we present a closed-form reconstruction algorithm for a special case of under-sampled MRI data. Usually, under-sampled MRI data requires iterative (instead of analytical) algorithms with L 1 -norm or total variation norm to reconstruct the image. (paper)

  7. A deep convolutional neural network for recognizing foods

    Science.gov (United States)

    Jahani Heravi, Elnaz; Habibi Aghdam, Hamed; Puig, Domenec

    2015-12-01

    Controlling the food intake is an efficient way that each person can undertake to tackle the obesity problem in countries worldwide. This is achievable by developing a smartphone application that is able to recognize foods and compute their calories. State-of-art methods are chiefly based on hand-crafted feature extraction methods such as HOG and Gabor. Recent advances in large-scale object recognition datasets such as ImageNet have revealed that deep Convolutional Neural Networks (CNN) possess more representation power than the hand-crafted features. The main challenge with CNNs is to find the appropriate architecture for each problem. In this paper, we propose a deep CNN which consists of 769; 988 parameters. Our experiments show that the proposed CNN outperforms the state-of-art methods and improves the best result of traditional methods 17%. Moreover, using an ensemble of two CNNs that have been trained two different times, we are able to improve the classification performance 21:5%.

  8. Forged Signature Distinction Using Convolutional Neural Network for Feature Extraction

    Directory of Open Access Journals (Sweden)

    Seungsoo Nam

    2018-01-01

    Full Text Available This paper proposes a dynamic verification scheme for finger-drawn signatures in smartphones. As a dynamic feature, the movement of a smartphone is recorded with accelerometer sensors in the smartphone, in addition to the moving coordinates of the signature. To extract high-level longitudinal and topological features, the proposed scheme uses a convolution neural network (CNN for feature extraction, and not as a conventional classifier. We assume that a CNN trained with forged signatures can extract effective features (called S-vector, which are common in forging activities such as hesitation and delay before drawing the complicated part. The proposed scheme also exploits an autoencoder (AE as a classifier, and the S-vector is used as the input vector to the AE. An AE has high accuracy for the one-class distinction problem such as signature verification, and is also greatly dependent on the accuracy of input data. S-vector is valuable as the input of AE, and, consequently, could lead to improved verification accuracy especially for distinguishing forged signatures. Compared to the previous work, i.e., the MLP-based finger-drawn signature verification scheme, the proposed scheme decreases the equal error rate by 13.7%, specifically, from 18.1% to 4.4%, for discriminating forged signatures.

  9. Convolutional networks for fast, energy-efficient neuromorphic computing.

    Science.gov (United States)

    Esser, Steven K; Merolla, Paul A; Arthur, John V; Cassidy, Andrew S; Appuswamy, Rathinakumar; Andreopoulos, Alexander; Berg, David J; McKinstry, Jeffrey L; Melano, Timothy; Barch, Davis R; di Nolfo, Carmelo; Datta, Pallab; Amir, Arnon; Taba, Brian; Flickner, Myron D; Modha, Dharmendra S

    2016-10-11

    Deep networks are now able to achieve human-level performance on a broad spectrum of recognition tasks. Independently, neuromorphic computing has now demonstrated unprecedented energy-efficiency through a new chip architecture based on spiking neurons, low precision synapses, and a scalable communication network. Here, we demonstrate that neuromorphic computing, despite its novel architectural primitives, can implement deep convolution networks that (i) approach state-of-the-art classification accuracy across eight standard datasets encompassing vision and speech, (ii) perform inference while preserving the hardware's underlying energy-efficiency and high throughput, running on the aforementioned datasets at between 1,200 and 2,600 frames/s and using between 25 and 275 mW (effectively >6,000 frames/s per Watt), and (iii) can be specified and trained using backpropagation with the same ease-of-use as contemporary deep learning. This approach allows the algorithmic power of deep learning to be merged with the efficiency of neuromorphic processors, bringing the promise of embedded, intelligent, brain-inspired computing one step closer.

  10. Convolution neural-network-based detection of lung structures

    Science.gov (United States)

    Hasegawa, Akira; Lo, Shih-Chung B.; Freedman, Matthew T.; Mun, Seong K.

    1994-05-01

    Chest radiography is one of the most primary and widely used techniques in diagnostic imaging. Nowadays with the advent of digital radiology, the digital medical image processing techniques for digital chest radiographs have attracted considerable attention, and several studies on the computer-aided diagnosis (CADx) as well as on the conventional image processing techniques for chest radiographs have been reported. In the automatic diagnostic process for chest radiographs, it is important to outline the areas of the lungs, the heart, and the diaphragm. This is because the original chest radiograph is composed of important anatomic structures and, without knowing exact positions of the organs, the automatic diagnosis may result in unexpected detections. The automatic extraction of an anatomical structure from digital chest radiographs can be a useful tool for (1) the evaluation of heart size, (2) automatic detection of interstitial lung diseases, (3) automatic detection of lung nodules, and (4) data compression, etc. Based on the clearly defined boundaries of heart area, rib spaces, rib positions, and rib cage extracted, one should be able to use this information to facilitate the tasks of the CADx on chest radiographs. In this paper, we present an automatic scheme for the detection of lung field from chest radiographs by using a shift-invariant convolution neural network. A novel algorithm for smoothing boundaries of lungs is also presented.

  11. Classification of crystal structure using a convolutional neural network.

    Science.gov (United States)

    Park, Woon Bae; Chung, Jiyong; Jung, Jaeyoung; Sohn, Keemin; Singh, Satendra Pal; Pyo, Myoungho; Shin, Namsoo; Sohn, Kee-Sun

    2017-07-01

    A deep machine-learning technique based on a convolutional neural network (CNN) is introduced. It has been used for the classification of powder X-ray diffraction (XRD) patterns in terms of crystal system, extinction group and space group. About 150 000 powder XRD patterns were collected and used as input for the CNN with no handcrafted engineering involved, and thereby an appropriate CNN architecture was obtained that allowed determination of the crystal system, extinction group and space group. In sharp contrast with the traditional use of powder XRD pattern analysis, the CNN never treats powder XRD patterns as a deconvoluted and discrete peak position or as intensity data, but instead the XRD patterns are regarded as nothing but a pattern similar to a picture. The CNN interprets features that humans cannot recognize in a powder XRD pattern. As a result, accuracy levels of 81.14, 83.83 and 94.99% were achieved for the space-group, extinction-group and crystal-system classifications, respectively. The well trained CNN was then used for symmetry identification of unknown novel inorganic compounds.

  12. Noisy Ocular Recognition Based on Three Convolutional Neural Networks

    Directory of Open Access Journals (Sweden)

    Min Beom Lee

    2017-12-01

    Full Text Available In recent years, the iris recognition system has been gaining increasing acceptance for applications such as access control and smartphone security. When the images of the iris are obtained under unconstrained conditions, an issue of undermined quality is caused by optical and motion blur, off-angle view (the user’s eyes looking somewhere else, not into the front of the camera, specular reflection (SR and other factors. Such noisy iris images increase intra-individual variations and, as a result, reduce the accuracy of iris recognition. A typical iris recognition system requires a near-infrared (NIR illuminator along with an NIR camera, which are larger and more expensive than fingerprint recognition equipment. Hence, many studies have proposed methods of using iris images captured by a visible light camera without the need for an additional illuminator. In this research, we propose a new recognition method for noisy iris and ocular images by using one iris and two periocular regions, based on three convolutional neural networks (CNNs. Experiments were conducted by using the noisy iris challenge evaluation-part II (NICE.II training dataset (selected from the university of Beira iris (UBIRIS.v2 database, mobile iris challenge evaluation (MICHE database, and institute of automation of Chinese academy of sciences (CASIA-Iris-Distance database. As a result, the method proposed by this study outperformed previous methods.

  13. Classification of breast cancer histology images using Convolutional Neural Networks.

    Directory of Open Access Journals (Sweden)

    Teresa Araújo

    Full Text Available Breast cancer is one of the main causes of cancer death worldwide. The diagnosis of biopsy tissue with hematoxylin and eosin stained images is non-trivial and specialists often disagree on the final diagnosis. Computer-aided Diagnosis systems contribute to reduce the cost and increase the efficiency of this process. Conventional classification approaches rely on feature extraction methods designed for a specific problem based on field-knowledge. To overcome the many difficulties of the feature-based approaches, deep learning methods are becoming important alternatives. A method for the classification of hematoxylin and eosin stained breast biopsy images using Convolutional Neural Networks (CNNs is proposed. Images are classified in four classes, normal tissue, benign lesion, in situ carcinoma and invasive carcinoma, and in two classes, carcinoma and non-carcinoma. The architecture of the network is designed to retrieve information at different scales, including both nuclei and overall tissue organization. This design allows the extension of the proposed system to whole-slide histology images. The features extracted by the CNN are also used for training a Support Vector Machine classifier. Accuracies of 77.8% for four class and 83.3% for carcinoma/non-carcinoma are achieved. The sensitivity of our method for cancer cases is 95.6%.

  14. Classification of Two Comic Books based on Convolutional Neural Networks

    Directory of Open Access Journals (Sweden)

    Miki UENO

    2017-03-01

    Full Text Available Unphotographic images are the powerful representations described various situations. Thus, understanding intellectual products such as comics and picture books is one of the important topics in the field of artificial intelligence. Hence, stepwise analysis of a comic story, i.e., features of a part of the image, information features, features relating to continuous scene etc., was pursued. Especially, the length and each scene of four-scene comics are limited so as to ensure a clear interpretation of the contents.In this study, as the first step in this direction, the problem to classify two four-scene comics by the same artists were focused as the example. Several classifiers were constructed by utilizing a Convolutional Neural Network(CNN, and the results of classification by a human annotator and by a computational method were compared.From these experiments, we have clearly shown that CNN is efficient way to classify unphotographic gray scaled images and found that characteristic features of images to classify incorrectly.

  15. Bone age detection via carpogram analysis using convolutional neural networks

    Science.gov (United States)

    Torres, Felipe; Bravo, María. Alejandra; Salinas, Emmanuel; Triana, Gustavo; Arbeláez, Pablo

    2017-11-01

    Bone age assessment is a critical factor for determining delayed development in children, which can be a sign of pathologies such as endocrine diseases, growth abnormalities, chromosomal, neurological and congenital disorders among others. In this paper we present BoneNet, a methodology to assess automatically the skeletal maturity state in pediatric patients based on Convolutional Neural Networks. We train and evaluate our algorithm on a database of X-Ray images provided by the hospital Fundacion Santa Fe de Bogot ´ a with around 1500 images of patients between the ages 1 to 18. ´ We compare two different architectures to classify the given data in order to explore the generality of our method. To accomplish this, we define multiple binary age assessment problems, dividing the data by bone age and differentiating the patients by their gender. Thus, exploring several parameters, we develop BoneNet. Our approach is holistic, efficient, and modular, since it is possible for the specialists to use all the networks combined to determine how is the skeletal maturity of a patient. BoneNet achieves over 90% accuracy for most of the critical age thresholds, when differentiating the images between over or under a given age.

  16. Automated embolic signal detection using Deep Convolutional Neural Network.

    Science.gov (United States)

    Sombune, Praotasna; Phienphanich, Phongphan; Phuechpanpaisal, Sutanya; Muengtaweepongsa, Sombat; Ruamthanthong, Anuchit; Tantibundhit, Charturong

    2017-07-01

    This work investigated the potential of Deep Neural Network in detection of cerebral embolic signal (ES) from transcranial Doppler ultrasound (TCD). The resulting system is aimed to couple with TCD devices in diagnosing a risk of stroke in real-time with high accuracy. The Adaptive Gain Control (AGC) approach developed in our previous study is employed to capture suspected ESs in real-time. By using spectrograms of the same TCD signal dataset as that of our previous work as inputs and the same experimental setup, Deep Convolutional Neural Network (CNN), which can learn features while training, was investigated for its ability to bypass the traditional handcrafted feature extraction and selection process. Extracted feature vectors from the suspected ESs are later determined whether they are of an ES, artifact (AF) or normal (NR) interval. The effectiveness of the developed system was evaluated over 19 subjects going under procedures generating emboli. The CNN-based system could achieve in average of 83.0% sensitivity, 80.1% specificity, and 81.4% accuracy, with considerably much less time consumption in development. The certainly growing set of training samples and computational resources will contribute to high performance. Besides having potential use in various clinical ES monitoring settings, continuation of this promising study will benefit developments of wearable applications by leveraging learnable features to serve demographic differentials.

  17. Deep convolutional neural network for mammographic density segmentation

    Science.gov (United States)

    Wei, Jun; Li, Songfeng; Chan, Heang-Ping; Helvie, Mark A.; Roubidoux, Marilyn A.; Lu, Yao; Zhou, Chuan; Hadjiiski, Lubomir; Samala, Ravi K.

    2018-02-01

    Breast density is one of the most significant factors for cancer risk. In this study, we proposed a supervised deep learning approach for automated estimation of percentage density (PD) on digital mammography (DM). The deep convolutional neural network (DCNN) was trained to estimate a probability map of breast density (PMD). PD was calculated as the ratio of the dense area to the breast area based on the probability of each pixel belonging to dense region or fatty region at a decision threshold of 0.5. The DCNN estimate was compared to a feature-based statistical learning approach, in which gray level, texture and morphological features were extracted from each ROI and the least absolute shrinkage and selection operator (LASSO) was used to select and combine the useful features to generate the PMD. The reference PD of each image was provided by two experienced MQSA radiologists. With IRB approval, we retrospectively collected 347 DMs from patient files at our institution. The 10-fold cross-validation results showed a strong correlation r=0.96 between the DCNN estimation and interactive segmentation by radiologists while that of the feature-based statistical learning approach vs radiologists' segmentation had a correlation r=0.78. The difference between the segmentation by DCNN and by radiologists was significantly smaller than that between the feature-based learning approach and radiologists (p approach has the potential to replace radiologists' interactive thresholding in PD estimation on DMs.

  18. Neonatal Seizure Detection Using Deep Convolutional Neural Networks.

    Science.gov (United States)

    Ansari, Amir H; Cherian, Perumpillichira J; Caicedo, Alexander; Naulaers, Gunnar; De Vos, Maarten; Van Huffel, Sabine

    2018-04-02

    Identifying a core set of features is one of the most important steps in the development of an automated seizure detector. In most of the published studies describing features and seizure classifiers, the features were hand-engineered, which may not be optimal. The main goal of the present paper is using deep convolutional neural networks (CNNs) and random forest to automatically optimize feature selection and classification. The input of the proposed classifier is raw multi-channel EEG and the output is the class label: seizure/nonseizure. By training this network, the required features are optimized, while fitting a nonlinear classifier on the features. After training the network with EEG recordings of 26 neonates, five end layers performing the classification were replaced with a random forest classifier in order to improve the performance. This resulted in a false alarm rate of 0.9 per hour and seizure detection rate of 77% using a test set of EEG recordings of 22 neonates that also included dubious seizures. The newly proposed CNN classifier outperformed three data-driven feature-based approaches and performed similar to a previously developed heuristic method.

  19. Cell dynamic morphology classification using deep convolutional neural networks.

    Science.gov (United States)

    Li, Heng; Pang, Fengqian; Shi, Yonggang; Liu, Zhiwen

    2018-05-15

    Cell morphology is often used as a proxy measurement of cell status to understand cell physiology. Hence, interpretation of cell dynamic morphology is a meaningful task in biomedical research. Inspired by the recent success of deep learning, we here explore the application of convolutional neural networks (CNNs) to cell dynamic morphology classification. An innovative strategy for the implementation of CNNs is introduced in this study. Mouse lymphocytes were collected to observe the dynamic morphology, and two datasets were thus set up to investigate the performances of CNNs. Considering the installation of deep learning, the classification problem was simplified from video data to image data, and was then solved by CNNs in a self-taught manner with the generated image data. CNNs were separately performed in three installation scenarios and compared with existing methods. Experimental results demonstrated the potential of CNNs in cell dynamic morphology classification, and validated the effectiveness of the proposed strategy. CNNs were successfully applied to the classification problem, and outperformed the existing methods in the classification accuracy. For the installation of CNNs, transfer learning was proved to be a promising scheme. © 2018 International Society for Advancement of Cytometry. © 2018 International Society for Advancement of Cytometry.

  20. Processing of chromatic information in a deep convolutional neural network.

    Science.gov (United States)

    Flachot, Alban; Gegenfurtner, Karl R

    2018-04-01

    Deep convolutional neural networks are a class of machine-learning algorithms capable of solving non-trivial tasks, such as object recognition, with human-like performance. Little is known about the exact computations that deep neural networks learn, and to what extent these computations are similar to the ones performed by the primate brain. Here, we investigate how color information is processed in the different layers of the AlexNet deep neural network, originally trained on object classification of over 1.2M images of objects in their natural contexts. We found that the color-responsive units in the first layer of AlexNet learned linear features and were broadly tuned to two directions in color space, analogously to what is known of color responsive cells in the primate thalamus. Moreover, these directions are decorrelated and lead to statistically efficient representations, similar to the cardinal directions of the second-stage color mechanisms in primates. We also found, in analogy to the early stages of the primate visual system, that chromatic and achromatic information were segregated in the early layers of the network. Units in the higher layers of AlexNet exhibit on average a lower responsivity for color than units at earlier stages.

  1. Noisy Ocular Recognition Based on Three Convolutional Neural Networks.

    Science.gov (United States)

    Lee, Min Beom; Hong, Hyung Gil; Park, Kang Ryoung

    2017-12-17

    In recent years, the iris recognition system has been gaining increasing acceptance for applications such as access control and smartphone security. When the images of the iris are obtained under unconstrained conditions, an issue of undermined quality is caused by optical and motion blur, off-angle view (the user's eyes looking somewhere else, not into the front of the camera), specular reflection (SR) and other factors. Such noisy iris images increase intra-individual variations and, as a result, reduce the accuracy of iris recognition. A typical iris recognition system requires a near-infrared (NIR) illuminator along with an NIR camera, which are larger and more expensive than fingerprint recognition equipment. Hence, many studies have proposed methods of using iris images captured by a visible light camera without the need for an additional illuminator. In this research, we propose a new recognition method for noisy iris and ocular images by using one iris and two periocular regions, based on three convolutional neural networks (CNNs). Experiments were conducted by using the noisy iris challenge evaluation-part II (NICE.II) training dataset (selected from the university of Beira iris (UBIRIS).v2 database), mobile iris challenge evaluation (MICHE) database, and institute of automation of Chinese academy of sciences (CASIA)-Iris-Distance database. As a result, the method proposed by this study outperformed previous methods.

  2. Understanding the Convolutional Neural Networks with Gradient Descent and Backpropagation

    Science.gov (United States)

    Zhou, XueFei

    2018-04-01

    With the development of computer technology, the applications of machine learning are more and more extensive. And machine learning is providing endless opportunities to develop new applications. One of those applications is image recognition by using Convolutional Neural Networks (CNNs). CNN is one of the most common algorithms in image recognition. It is significant to understand its theory and structure for every scholar who is interested in this field. CNN is mainly used in computer identification, especially in voice, text recognition and other aspects of the application. It utilizes hierarchical structure with different layers to accelerate computing speed. In addition, the greatest features of CNNs are the weight sharing and dimension reduction. And all of these consolidate the high effectiveness and efficiency of CNNs with idea computing speed and error rate. With the help of other learning altruisms, CNNs could be used in several scenarios for machine learning, especially for deep learning. Based on the general introduction to the background and the core solution CNN, this paper is going to focus on summarizing how Gradient Descent and Backpropagation work, and how they contribute to the high performances of CNNs. Also, some practical applications will be discussed in the following parts. The last section exhibits the conclusion and some perspectives of future work.

  3. Convolutional networks for fast, energy-efficient neuromorphic computing

    Science.gov (United States)

    Esser, Steven K.; Merolla, Paul A.; Arthur, John V.; Cassidy, Andrew S.; Appuswamy, Rathinakumar; Andreopoulos, Alexander; Berg, David J.; McKinstry, Jeffrey L.; Melano, Timothy; Barch, Davis R.; di Nolfo, Carmelo; Datta, Pallab; Amir, Arnon; Taba, Brian; Flickner, Myron D.; Modha, Dharmendra S.

    2016-01-01

    Deep networks are now able to achieve human-level performance on a broad spectrum of recognition tasks. Independently, neuromorphic computing has now demonstrated unprecedented energy-efficiency through a new chip architecture based on spiking neurons, low precision synapses, and a scalable communication network. Here, we demonstrate that neuromorphic computing, despite its novel architectural primitives, can implement deep convolution networks that (i) approach state-of-the-art classification accuracy across eight standard datasets encompassing vision and speech, (ii) perform inference while preserving the hardware’s underlying energy-efficiency and high throughput, running on the aforementioned datasets at between 1,200 and 2,600 frames/s and using between 25 and 275 mW (effectively >6,000 frames/s per Watt), and (iii) can be specified and trained using backpropagation with the same ease-of-use as contemporary deep learning. This approach allows the algorithmic power of deep learning to be merged with the efficiency of neuromorphic processors, bringing the promise of embedded, intelligent, brain-inspired computing one step closer. PMID:27651489

  4. Axillary Lymph Node Evaluation Utilizing Convolutional Neural Networks Using MRI Dataset.

    Science.gov (United States)

    Ha, Richard; Chang, Peter; Karcich, Jenika; Mutasa, Simukayi; Fardanesh, Reza; Wynn, Ralph T; Liu, Michael Z; Jambawalikar, Sachin

    2018-04-25

    The aim of this study is to evaluate the role of convolutional neural network (CNN) in predicting axillary lymph node metastasis, using a breast MRI dataset. An institutional review board (IRB)-approved retrospective review of our database from 1/2013 to 6/2016 identified 275 axillary lymph nodes for this study. Biopsy-proven 133 metastatic axillary lymph nodes and 142 negative control lymph nodes were identified based on benign biopsies (100) and from healthy MRI screening patients (42) with at least 3 years of negative follow-up. For each breast MRI, axillary lymph node was identified on first T1 post contrast dynamic images and underwent 3D segmentation using an open source software platform 3D Slicer. A 32 × 32 patch was then extracted from the center slice of the segmented tumor data. A CNN was designed for lymph node prediction based on each of these cropped images. The CNN consisted of seven convolutional layers and max-pooling layers with 50% dropout applied in the linear layer. In addition, data augmentation and L2 regularization were performed to limit overfitting. Training was implemented using the Adam optimizer, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments. Code for this study was written in Python using the TensorFlow module (1.0.0). Experiments and CNN training were done on a Linux workstation with NVIDIA GTX 1070 Pascal GPU. Two class axillary lymph node metastasis prediction models were evaluated. For each lymph node, a final softmax score threshold of 0.5 was used for classification. Based on this, CNN achieved a mean five-fold cross-validation accuracy of 84.3%. It is feasible for current deep CNN architectures to be trained to predict likelihood of axillary lymph node metastasis. Larger dataset will likely improve our prediction model and can potentially be a non-invasive alternative to core needle biopsy and even sentinel lymph node

  5. Optical transmission testing based on asynchronous sampling techniques: images analysis containing chromatic dispersion using convolutional neural network

    Science.gov (United States)

    Mrozek, T.; Perlicki, K.; Tajmajer, T.; Wasilewski, P.

    2017-08-01

    The article presents an image analysis method, obtained from an asynchronous delay tap sampling (ADTS) technique, which is used for simultaneous monitoring of various impairments occurring in the physical layer of the optical network. The ADTS method enables the visualization of the optical signal in the form of characteristics (so called phase portraits) that change their shape under the influence of impairments such as chromatic dispersion, polarization mode dispersion and ASE noise. Using this method, a simulation model was built with OptSim 4.0. After the simulation study, data were obtained in the form of images that were further analyzed using the convolutional neural network algorithm. The main goal of the study was to train a convolutional neural network to recognize the selected impairment (distortion); then to test its accuracy and estimate the impairment for the selected set of test images. The input data consisted of processed binary images in the form of two-dimensional matrices, with the position of the pixel. This article focuses only on the analysis of images containing chromatic dispersion.

  6. Remote Sensing Scene Classification Based on Convolutional Neural Networks Pre-Trained Using Attention-Guided Sparse Filters

    Directory of Open Access Journals (Sweden)

    Jingbo Chen

    2018-02-01

    Full Text Available Semantic-level land-use scene classification is a challenging problem, in which deep learning methods, e.g., convolutional neural networks (CNNs, have shown remarkable capacity. However, a lack of sufficient labeled images has proved a hindrance to increasing the land-use scene classification accuracy of CNNs. Aiming at this problem, this paper proposes a CNN pre-training method under the guidance of a human visual attention mechanism. Specifically, a computational visual attention model is used to automatically extract salient regions in unlabeled images. Then, sparse filters are adopted to learn features from these salient regions, with the learnt parameters used to initialize the convolutional layers of the CNN. Finally, the CNN is further fine-tuned on labeled images. Experiments are performed on the UCMerced and AID datasets, which show that when combined with a demonstrative CNN, our method can achieve 2.24% higher accuracy than a plain CNN and can obtain an overall accuracy of 92.43% when combined with AlexNet. The results indicate that the proposed method can effectively improve CNN performance using easy-to-access unlabeled images and thus will enhance the performance of land-use scene classification especially when a large-scale labeled dataset is unavailable.

  7. Deep 3D Convolutional Encoder Networks With Shortcuts for Multiscale Feature Integration Applied to Multiple Sclerosis Lesion Segmentation.

    Science.gov (United States)

    Brosch, Tom; Tang, Lisa Y W; Youngjin Yoo; Li, David K B; Traboulsee, Anthony; Tam, Roger

    2016-05-01

    We propose a novel segmentation approach based on deep 3D convolutional encoder networks with shortcut connections and apply it to the segmentation of multiple sclerosis (MS) lesions in magnetic resonance images. Our model is a neural network that consists of two interconnected pathways, a convolutional pathway, which learns increasingly more abstract and higher-level image features, and a deconvolutional pathway, which predicts the final segmentation at the voxel level. The joint training of the feature extraction and prediction pathways allows for the automatic learning of features at different scales that are optimized for accuracy for any given combination of image types and segmentation task. In addition, shortcut connections between the two pathways allow high- and low-level features to be integrated, which enables the segmentation of lesions across a wide range of sizes. We have evaluated our method on two publicly available data sets (MICCAI 2008 and ISBI 2015 challenges) with the results showing that our method performs comparably to the top-ranked state-of-the-art methods, even when only relatively small data sets are available for training. In addition, we have compared our method with five freely available and widely used MS lesion segmentation methods (EMS, LST-LPA, LST-LGA, Lesion-TOADS, and SLS) on a large data set from an MS clinical trial. The results show that our method consistently outperforms these other methods across a wide range of lesion sizes.

  8. Data-based diffraction kernels for surface waves from convolution and correlation processes through active seismic interferometry

    Science.gov (United States)

    Chmiel, Malgorzata; Roux, Philippe; Herrmann, Philippe; Rondeleux, Baptiste; Wathelet, Marc

    2018-05-01

    We investigated the construction of diffraction kernels for surface waves using two-point convolution and/or correlation from land active seismic data recorded in the context of exploration geophysics. The high density of controlled sources and receivers, combined with the application of the reciprocity principle, allows us to retrieve two-dimensional phase-oscillation diffraction kernels (DKs) of surface waves between any two source or receiver points in the medium at each frequency (up to 15 Hz, at least). These DKs are purely data-based as no model calculations and no synthetic data are needed. They naturally emerge from the interference patterns of the recorded wavefields projected on the dense array of sources and/or receivers. The DKs are used to obtain multi-mode dispersion relations of Rayleigh waves, from which near-surface shear velocity can be extracted. Using convolution versus correlation with a grid of active sources is an important step in understanding the physics of the retrieval of surface wave Green's functions. This provides the foundation for future studies based on noise sources or active sources with a sparse spatial distribution.

  9. Semantic labeling of high-resolution aerial images using an ensemble of fully convolutional networks

    Science.gov (United States)

    Sun, Xiaofeng; Shen, Shuhan; Lin, Xiangguo; Hu, Zhanyi

    2017-10-01

    High-resolution remote sensing data classification has been a challenging and promising research topic in the community of remote sensing. In recent years, with the rapid advances of deep learning, remarkable progress has been made in this field, which facilitates a transition from hand-crafted features designing to an automatic end-to-end learning. A deep fully convolutional networks (FCNs) based ensemble learning method is proposed to label the high-resolution aerial images. To fully tap the potentials of FCNs, both the Visual Geometry Group network and a deeper residual network, ResNet, are employed. Furthermore, to enlarge training samples with diversity and gain better generalization, in addition to the commonly used data augmentation methods (e.g., rotation, multiscale, and aspect ratio) in the literature, aerial images from other datasets are also collected for cross-scene learning. Finally, we combine these learned models to form an effective FCN ensemble and refine the results using a fully connected conditional random field graph model. Experiments on the ISPRS 2-D Semantic Labeling Contest dataset show that our proposed end-to-end classification method achieves an overall accuracy of 90.7%, a state-of-the-art in the field.

  10. Classifying the molecular functions of Rab GTPases in membrane trafficking using deep convolutional neural networks.

    Science.gov (United States)

    Le, Nguyen-Quoc-Khanh; Ho, Quang-Thai; Ou, Yu-Yen

    2018-06-13

    Deep learning has been increasingly used to solve a number of problems with state-of-the-art performance in a wide variety of fields. In biology, deep learning can be applied to reduce feature extraction time and achieve high levels of performance. In our present work, we apply deep learning via two-dimensional convolutional neural networks and position-specific scoring matrices to classify Rab protein molecules, which are main regulators in membrane trafficking for transferring proteins and other macromolecules throughout the cell. The functional loss of specific Rab molecular functions has been implicated in a variety of human diseases, e.g., choroideremia, intellectual disabilities, cancer. Therefore, creating a precise model for classifying Rabs is crucial in helping biologists understand the molecular functions of Rabs and design drug targets according to such specific human disease information. We constructed a robust deep neural network for classifying Rabs that achieved an accuracy of 99%, 99.5%, 96.3%, and 97.6% for each of four specific molecular functions. Our approach demonstrates superior performance to traditional artificial neural networks. Therefore, from our proposed study, we provide both an effective tool for classifying Rab proteins and a basis for further research that can improve the performance of biological modeling using deep neural networks. Copyright © 2018 Elsevier Inc. All rights reserved.

  11. Convolutional neural networks for transient candidate vetting in large-scale surveys

    Science.gov (United States)

    Gieseke, Fabian; Bloemen, Steven; van den Bogaard, Cas; Heskes, Tom; Kindler, Jonas; Scalzo, Richard A.; Ribeiro, Valério A. R. M.; van Roestel, Jan; Groot, Paul J.; Yuan, Fang; Möller, Anais; Tucker, Brad E.

    2017-12-01

    Current synoptic sky surveys monitor large areas of the sky to find variable and transient astronomical sources. As the number of detections per night at a single telescope easily exceeds several thousand, current detection pipelines make intensive use of machine learning algorithms to classify the detected objects and to filter out the most interesting candidates. A number of upcoming surveys will produce up to three orders of magnitude more data, which renders high-precision classification systems essential to reduce the manual and, hence, expensive vetting by human experts. We present an approach based on convolutional neural networks to discriminate between true astrophysical sources and artefacts in reference-subtracted optical images. We show that relatively simple networks are already competitive with state-of-the-art systems and that their quality can further be improved via slightly deeper networks and additional pre-processing steps - eventually yielding models outperforming state-of-the-art systems. In particular, our best model correctly classifies about 97.3 per cent of all 'real' and 99.7 per cent of all 'bogus' instances on a test set containing 1942 'bogus' and 227 'real' instances in total. Furthermore, the networks considered in this work can also successfully classify these objects at hand without relying on difference images, which might pave the way for future detection pipelines not containing image subtraction steps at all.

  12. Automated EEG-based screening of depression using deep convolutional neural network.

    Science.gov (United States)

    Acharya, U Rajendra; Oh, Shu Lih; Hagiwara, Yuki; Tan, Jen Hong; Adeli, Hojjat; Subha, D P

    2018-07-01

    In recent years, advanced neurocomputing and machine learning techniques have been used for Electroencephalogram (EEG)-based diagnosis of various neurological disorders. In this paper, a novel computer model is presented for EEG-based screening of depression using a deep neural network machine learning approach, known as Convolutional Neural Network (CNN). The proposed technique does not require a semi-manually-selected set of features to be fed into a classifier for classification. It learns automatically and adaptively from the input EEG signals to differentiate EEGs obtained from depressive and normal subjects. The model was tested using EEGs obtained from 15 normal and 15 depressed patients. The algorithm attained accuracies of 93.5% and 96.0% using EEG signals from the left and right hemisphere, respectively. It was discovered in this research that the EEG signals from the right hemisphere are more distinctive in depression than those from the left hemisphere. This discovery is consistent with recent research and revelation that the depression is associated with a hyperactive right hemisphere. An exciting extension of this research would be diagnosis of different stages and severity of depression and development of a Depression Severity Index (DSI). Copyright © 2018 Elsevier B.V. All rights reserved.

  13. A method for medulloblastoma tumor differentiation based on convolutional neural networks and transfer learning

    Science.gov (United States)

    Cruz-Roa, Angel; Arévalo, John; Judkins, Alexander; Madabhushi, Anant; González, Fabio

    2015-12-01

    Convolutional neural networks (CNN) have been very successful at addressing different computer vision tasks thanks to their ability to learn image representations directly from large amounts of labeled data. Features learned from a dataset can be used to represent images from a different dataset via an approach called transfer learning. In this paper we apply transfer learning to the challenging task of medulloblastoma tumor differentiation. We compare two different CNN models which were previously trained in two different domains (natural and histopathology images). The first CNN is a state-of-the-art approach in computer vision, a large and deep CNN with 16-layers, Visual Geometry Group (VGG) CNN. The second (IBCa-CNN) is a 2-layer CNN trained for invasive breast cancer tumor classification. Both CNNs are used as visual feature extractors of histopathology image regions of anaplastic and non-anaplastic medulloblastoma tumor from digitized whole-slide images. The features from the two models are used, separately, to train a softmax classifier to discriminate between anaplastic and non-anaplastic medulloblastoma image regions. Experimental results show that the transfer learning approach produce competitive results in comparison with the state of the art approaches for IBCa detection. Results also show that features extracted from the IBCa-CNN have better performance in comparison with features extracted from the VGG-CNN. The former obtains 89.8% while the latter obtains 76.6% in terms of average accuracy.

  14. Breast mass detection in mammography and tomosynthesis via fully convolutional network-based heatmap regression

    Science.gov (United States)

    Zhang, Jun; Cain, Elizabeth Hope; Saha, Ashirbani; Zhu, Zhe; Mazurowski, Maciej A.

    2018-02-01

    Breast mass detection in mammography and digital breast tomosynthesis (DBT) is an essential step in computerized breast cancer analysis. Deep learning-based methods incorporate feature extraction and model learning into a unified framework and have achieved impressive performance in various medical applications (e.g., disease diagnosis, tumor detection, and landmark detection). However, these methods require large-scale accurately annotated data. Unfortunately, it is challenging to get precise annotations of breast masses. To address this issue, we propose a fully convolutional network (FCN) based heatmap regression method for breast mass detection, using only weakly annotated mass regions in mammography images. Specifically, we first generate heat maps of masses based on human-annotated rough regions for breast masses. We then develop an FCN model for end-to-end heatmap regression with an F-score loss function, where the mammography images are regarded as the input and heatmaps for breast masses are used as the output. Finally, the probability map of mass locations can be estimated with the trained model. Experimental results on a mammography dataset with 439 subjects demonstrate the effectiveness of our method. Furthermore, we evaluate whether we can use mammography data to improve detection models for DBT, since mammography shares similar structure with tomosynthesis. We propose a transfer learning strategy by fine-tuning the learned FCN model from mammography images. We test this approach on a small tomosynthesis dataset with only 40 subjects, and we show an improvement in the detection performance as compared to training the model from scratch.

  15. Two-stream Convolutional Neural Network for Methane Emissions Quantification

    Science.gov (United States)

    Wang, J.; Ravikumar, A. P.; McGuire, M.; Bell, C.; Tchapmi, L. P.; Brandt, A. R.

    2017-12-01

    Methane, a key component of natural gas, has a 25x higher global warming potential than carbon dioxide on a 100-year basis. Accurately monitoring and mitigating methane emissions require cost-effective detection and quantification technologies. Optical gas imaging, one of the most commonly used leak detection technology, adopted by Environmental Protection Agency, cannot estimate leak-sizes. In this work, we harness advances in computer science to allow for rapid and automatic leak quantification. Particularly, we utilize two-stream deep Convolutional Networks (ConvNets) to estimate leak-size by capturing complementary spatial information from still plume frames, and temporal information from plume motion between frames. We build large leak datasets for training and evaluating purposes by collecting about 20 videos (i.e. 397,400 frames) of leaks. The videos were recorded at six distances from the source, covering 10 -60 ft. Leak sources included natural gas well-heads, separators, and tanks. All frames were labeled with a true leak size, which has eight levels ranging from 0 to 140 MCFH. Preliminary analysis shows that two-stream ConvNets provides significant accuracy advantage over single steam ConvNets. Spatial stream ConvNet can achieve an accuracy of 65.2%, by extracting important features, including texture, plume area, and pattern. Temporal stream, fed by the results of optical flow analysis, results in an accuracy of 58.3%. The integration of the two-stream ConvNets gives a combined accuracy of 77.6%. For future work, we will split the training and testing datasets in distinct ways in order to test the generalization of the algorithm for different leak sources. Several analytic metrics, including confusion matrix and visualization of key features, will be used to understand accuracy rates and occurrences of false positives. The quantification algorithm can help to find and fix super-emitters, and improve the cost-effectiveness of leak detection and repair

  16. A convolution-superposition dose calculation engine for GPUs

    Energy Technology Data Exchange (ETDEWEB)

    Hissoiny, Sami; Ozell, Benoit; Despres, Philippe [Departement de genie informatique et genie logiciel, Ecole polytechnique de Montreal, 2500 Chemin de Polytechnique, Montreal, Quebec H3T 1J4 (Canada); Departement de radio-oncologie, CRCHUM-Centre hospitalier de l' Universite de Montreal, 1560 rue Sherbrooke Est, Montreal, Quebec H2L 4M1 (Canada)

    2010-03-15

    Purpose: Graphic processing units (GPUs) are increasingly used for scientific applications, where their parallel architecture and unprecedented computing power density can be exploited to accelerate calculations. In this paper, a new GPU implementation of a convolution/superposition (CS) algorithm is presented. Methods: This new GPU implementation has been designed from the ground-up to use the graphics card's strengths and to avoid its weaknesses. The CS GPU algorithm takes into account beam hardening, off-axis softening, kernel tilting, and relies heavily on raytracing through patient imaging data. Implementation details are reported as well as a multi-GPU solution. Results: An overall single-GPU acceleration factor of 908x was achieved when compared to a nonoptimized version of the CS algorithm implemented in PlanUNC in single threaded central processing unit (CPU) mode, resulting in approximatively 2.8 s per beam for a 3D dose computation on a 0.4 cm grid. A comparison to an established commercial system leads to an acceleration factor of approximately 29x or 0.58 versus 16.6 s per beam in single threaded mode. An acceleration factor of 46x has been obtained for the total energy released per mass (TERMA) calculation and a 943x acceleration factor for the CS calculation compared to PlanUNC. Dose distributions also have been obtained for a simple water-lung phantom to verify that the implementation gives accurate results. Conclusions: These results suggest that GPUs are an attractive solution for radiation therapy applications and that careful design, taking the GPU architecture into account, is critical in obtaining significant acceleration factors. These results potentially can have a significant impact on complex dose delivery techniques requiring intensive dose calculations such as intensity-modulated radiation therapy (IMRT) and arc therapy. They also are relevant for adaptive radiation therapy where dose results must be obtained rapidly.

  17. Automated segmentation of geographic atrophy using deep convolutional neural networks

    Science.gov (United States)

    Hu, Zhihong; Wang, Ziyuan; Sadda, SriniVas R.

    2018-02-01

    Geographic atrophy (GA) is an end-stage manifestation of the advanced age-related macular degeneration (AMD), the leading cause of blindness and visual impairment in developed nations. Techniques to rapidly and precisely detect and quantify GA would appear to be of critical importance in advancing the understanding of its pathogenesis. In this study, we develop an automated supervised classification system using deep convolutional neural networks (CNNs) for segmenting GA in fundus autofluorescene (FAF) images. More specifically, to enhance the contrast of GA relative to the background, we apply the contrast limited adaptive histogram equalization. Blood vessels may cause GA segmentation errors due to similar intensity level to GA. A tensor-voting technique is performed to identify the blood vessels and a vessel inpainting technique is applied to suppress the GA segmentation errors due to the blood vessels. To handle the large variation of GA lesion sizes, three deep CNNs with three varying sized input image patches are applied. Fifty randomly chosen FAF images are obtained from fifty subjects with GA. The algorithm-defined GA regions are compared with manual delineation by a certified grader. A two-fold cross-validation is applied to evaluate the algorithm performance. The mean segmentation accuracy, true positive rate (i.e. sensitivity), true negative rate (i.e. specificity), positive predictive value, false discovery rate, and overlap ratio, between the algorithm- and manually-defined GA regions are 0.97 +/- 0.02, 0.89 +/- 0.08, 0.98 +/- 0.02, 0.87 +/- 0.12, 0.13 +/- 0.12, and 0.79 +/- 0.12 respectively, demonstrating a high level of agreement.

  18. Aerial Images and Convolutional Neural Network for Cotton Bloom Detection.

    Science.gov (United States)

    Xu, Rui; Li, Changying; Paterson, Andrew H; Jiang, Yu; Sun, Shangpeng; Robertson, Jon S

    2017-01-01

    Monitoring flower development can provide useful information for production management, estimating yield and selecting specific genotypes of crops. The main goal of this study was to develop a methodology to detect and count cotton flowers, or blooms, using color images acquired by an unmanned aerial system. The aerial images were collected from two test fields in 4 days. A convolutional neural network (CNN) was designed and trained to detect cotton blooms in raw images, and their 3D locations were calculated using the dense point cloud constructed from the aerial images with the structure from motion method. The quality of the dense point cloud was analyzed and plots with poor quality were excluded from data analysis. A constrained clustering algorithm was developed to register the same bloom detected from different images based on the 3D location of the bloom. The accuracy and incompleteness of the dense point cloud were analyzed because they affected the accuracy of the 3D location of the blooms and thus the accuracy of the bloom registration result. The constrained clustering algorithm was validated using simulated data, showing good efficiency and accuracy. The bloom count from the proposed method was comparable with the number counted manually with an error of -4 to 3 blooms for the field with a single plant per plot. However, more plots were underestimated in the field with multiple plants per plot due to hidden blooms that were not captured by the aerial images. The proposed methodology provides a high-throughput method to continuously monitor the flowering progress of cotton.

  19. Classifying Radio Galaxies with the Convolutional Neural Network

    International Nuclear Information System (INIS)

    Aniyan, A. K.; Thorat, K.

    2017-01-01

    We present the application of a deep machine learning technique to classify radio images of extended sources on a morphological basis using convolutional neural networks (CNN). In this study, we have taken the case of the Fanaroff–Riley (FR) class of radio galaxies as well as radio galaxies with bent-tailed morphology. We have used archival data from the Very Large Array (VLA)—Faint Images of the Radio Sky at Twenty Centimeters survey and existing visually classified samples available in the literature to train a neural network for morphological classification of these categories of radio sources. Our training sample size for each of these categories is ∼200 sources, which has been augmented by rotated versions of the same. Our study shows that CNNs can classify images of the FRI and FRII and bent-tailed radio galaxies with high accuracy (maximum precision at 95%) using well-defined samples and a “fusion classifier,” which combines the results of binary classifications, while allowing for a mechanism to find sources with unusual morphologies. The individual precision is highest for bent-tailed radio galaxies at 95% and is 91% and 75% for the FRI and FRII classes, respectively, whereas the recall is highest for FRI and FRIIs at 91% each, while the bent-tailed class has a recall of 79%. These results show that our results are comparable to that of manual classification, while being much faster. Finally, we discuss the computational and data-related challenges associated with the morphological classification of radio galaxies with CNNs.

  20. Convolutional neural networks for prostate cancer recurrence prediction

    Science.gov (United States)

    Kumar, Neeraj; Verma, Ruchika; Arora, Ashish; Kumar, Abhay; Gupta, Sanchit; Sethi, Amit; Gann, Peter H.

    2017-03-01

    Accurate prediction of the treatment outcome is important for cancer treatment planning. We present an approach to predict prostate cancer (PCa) recurrence after radical prostatectomy using tissue images. We used a cohort whose case vs. control (recurrent vs. non-recurrent) status had been determined using post-treatment follow up. Further, to aid the development of novel biomarkers of PCa recurrence, cases and controls were paired based on matching of other predictive clinical variables such as Gleason grade, stage, age, and race. For this cohort, tissue resection microarray with up to four cores per patient was available. The proposed approach is based on deep learning, and its novelty lies in the use of two separate convolutional neural networks (CNNs) - one to detect individual nuclei even in the crowded areas, and the other to classify them. To detect nuclear centers in an image, the first CNN predicts distance transform of the underlying (but unknown) multi-nuclear map from the input HE image. The second CNN classifies the patches centered at nuclear centers into those belonging to cases or controls. Voting across patches extracted from image(s) of a patient yields the probability of recurrence for the patient. The proposed approach gave 0.81 AUC for a sample of 30 recurrent cases and 30 non-recurrent controls, after being trained on an independent set of 80 case-controls pairs. If validated further, such an approach might help in choosing between a combination of treatment options such as active surveillance, radical prostatectomy, radiation, and hormone therapy. It can also generalize to the prediction of treatment outcomes in other cancers.

  1. Classifying Radio Galaxies with the Convolutional Neural Network

    Energy Technology Data Exchange (ETDEWEB)

    Aniyan, A. K.; Thorat, K. [Department of Physics and Electronics, Rhodes University, Grahamstown (South Africa)

    2017-06-01

    We present the application of a deep machine learning technique to classify radio images of extended sources on a morphological basis using convolutional neural networks (CNN). In this study, we have taken the case of the Fanaroff–Riley (FR) class of radio galaxies as well as radio galaxies with bent-tailed morphology. We have used archival data from the Very Large Array (VLA)—Faint Images of the Radio Sky at Twenty Centimeters survey and existing visually classified samples available in the literature to train a neural network for morphological classification of these categories of radio sources. Our training sample size for each of these categories is ∼200 sources, which has been augmented by rotated versions of the same. Our study shows that CNNs can classify images of the FRI and FRII and bent-tailed radio galaxies with high accuracy (maximum precision at 95%) using well-defined samples and a “fusion classifier,” which combines the results of binary classifications, while allowing for a mechanism to find sources with unusual morphologies. The individual precision is highest for bent-tailed radio galaxies at 95% and is 91% and 75% for the FRI and FRII classes, respectively, whereas the recall is highest for FRI and FRIIs at 91% each, while the bent-tailed class has a recall of 79%. These results show that our results are comparable to that of manual classification, while being much faster. Finally, we discuss the computational and data-related challenges associated with the morphological classification of radio galaxies with CNNs.

  2. Classifying Radio Galaxies with the Convolutional Neural Network

    Science.gov (United States)

    Aniyan, A. K.; Thorat, K.

    2017-06-01

    We present the application of a deep machine learning technique to classify radio images of extended sources on a morphological basis using convolutional neural networks (CNN). In this study, we have taken the case of the Fanaroff-Riley (FR) class of radio galaxies as well as radio galaxies with bent-tailed morphology. We have used archival data from the Very Large Array (VLA)—Faint Images of the Radio Sky at Twenty Centimeters survey and existing visually classified samples available in the literature to train a neural network for morphological classification of these categories of radio sources. Our training sample size for each of these categories is ˜200 sources, which has been augmented by rotated versions of the same. Our study shows that CNNs can classify images of the FRI and FRII and bent-tailed radio galaxies with high accuracy (maximum precision at 95%) using well-defined samples and a “fusion classifier,” which combines the results of binary classifications, while allowing for a mechanism to find sources with unusual morphologies. The individual precision is highest for bent-tailed radio galaxies at 95% and is 91% and 75% for the FRI and FRII classes, respectively, whereas the recall is highest for FRI and FRIIs at 91% each, while the bent-tailed class has a recall of 79%. These results show that our results are comparable to that of manual classification, while being much faster. Finally, we discuss the computational and data-related challenges associated with the morphological classification of radio galaxies with CNNs.

  3. Deep convolutional networks for pancreas segmentation in CT imaging

    Science.gov (United States)

    Roth, Holger R.; Farag, Amal; Lu, Le; Turkbey, Evrim B.; Summers, Ronald M.

    2015-03-01

    Automatic organ segmentation is an important prerequisite for many computer-aided diagnosis systems. The high anatomical variability of organs in the abdomen, such as the pancreas, prevents many segmentation methods from achieving high accuracies when compared to state-of-the-art segmentation of organs like the liver, heart or kidneys. Recently, the availability of large annotated training sets and the accessibility of affordable parallel computing resources via GPUs have made it feasible for "deep learning" methods such as convolutional networks (ConvNets) to succeed in image classification tasks. These methods have the advantage that used classification features are trained directly from the imaging data. We present a fully-automated bottom-up method for pancreas segmentation in computed tomography (CT) images of the abdomen. The method is based on hierarchical coarse-to-fine classification of local image regions (superpixels). Superpixels are extracted from the abdominal region using Simple Linear Iterative Clustering (SLIC). An initial probability response map is generated, using patch-level confidences and a two-level cascade of random forest classifiers, from which superpixel regions with probabilities larger 0.5 are retained. These retained superpixels serve as a highly sensitive initial input of the pancreas and its surroundings to a ConvNet that samples a bounding box around each superpixel at different scales (and random non-rigid deformations at training time) in order to assign a more distinct probability of each superpixel region being pancreas or not. We evaluate our method on CT images of 82 patients (60 for training, 2 for validation, and 20 for testing). Using ConvNets we achieve maximum Dice scores of an average 68% +/- 10% (range, 43-80%) in testing. This shows promise for accurate pancreas segmentation, using a deep learning approach and compares favorably to state-of-the-art methods.

  4. Contextual Multi-Scale Region Convolutional 3D Network for Activity Detection

    KAUST Repository

    Bai, Yancheng

    2018-01-28

    Activity detection is a fundamental problem in computer vision. Detecting activities of different temporal scales is particularly challenging. In this paper, we propose the contextual multi-scale region convolutional 3D network (CMS-RC3D) for activity detection. To deal with the inherent temporal scale variability of activity instances, the temporal feature pyramid is used to represent activities of different temporal scales. On each level of the temporal feature pyramid, an activity proposal detector and an activity classifier are learned to detect activities of specific temporal scales. Temporal contextual information is fused into activity classifiers for better recognition. More importantly, the entire model at all levels can be trained end-to-end. Our CMS-RC3D detector can deal with activities at all temporal scale ranges with only a single pass through the backbone network. We test our detector on two public activity detection benchmarks, THUMOS14 and ActivityNet. Extensive experiments show that the proposed CMS-RC3D detector outperforms state-of-the-art methods on THUMOS14 by a substantial margin and achieves comparable results on ActivityNet despite using a shallow feature extractor.

  5. Convolutional neural network-based classification system design with compressed wireless sensor network images.

    Science.gov (United States)

    Ahn, Jungmo; Park, JaeYeon; Park, Donghwan; Paek, Jeongyeup; Ko, JeongGil

    2018-01-01

    With the introduction of various advanced deep learning algorithms, initiatives for image classification systems have transitioned over from traditional machine learning algorithms (e.g., SVM) to Convolutional Neural Networks (CNNs) using deep learning software tools. A prerequisite in applying CNN to real world applications is a system that collects meaningful and useful data. For such purposes, Wireless Image Sensor Networks (WISNs), that are capable of monitoring natural environment phenomena using tiny and low-power cameras on resource-limited embedded devices, can be considered as an effective means of data collection. However, with limited battery resources, sending high-resolution raw images to the backend server is a burdensome task that has direct impact on network lifetime. To address this problem, we propose an energy-efficient pre- and post- processing mechanism using image resizing and color quantization that can significantly reduce the amount of data transferred while maintaining the classification accuracy in the CNN at the backend server. We show that, if well designed, an image in its highly compressed form can be well-classified with a CNN model trained in advance using adequately compressed data. Our evaluation using a real image dataset shows that an embedded device can reduce the amount of transmitted data by ∼71% while maintaining a classification accuracy of ∼98%. Under the same conditions, this process naturally reduces energy consumption by ∼71% compared to a WISN that sends the original uncompressed images.

  6. Eigennoise Speech Recovery in Adverse Environments with Joint Compensation of Additive and Convolutive Noise

    Directory of Open Access Journals (Sweden)

    Trung-Nghia Phung

    2015-01-01

    Full Text Available The learning-based speech recovery approach using statistical spectral conversion has been used for some kind of distorted speech as alaryngeal speech and body-conducted speech (or bone-conducted speech. This approach attempts to recover clean speech (undistorted speech from noisy speech (distorted speech by converting the statistical models of noisy speech into that of clean speech without the prior knowledge on characteristics and distributions of noise source. Presently, this approach has still not attracted many researchers to apply in general noisy speech enhancement because of some major problems: those are the difficulties of noise adaptation and the lack of noise robust synthesizable features in different noisy environments. In this paper, we adopted the methods of state-of-the-art voice conversions and speaker adaptation in speech recognition to the proposed speech recovery approach applied in different kinds of noisy environment, especially in adverse environments with joint compensation of additive and convolutive noises. We proposed to use the decorrelated wavelet packet coefficients as a low-dimensional robust synthesizable feature under noisy environments. We also proposed a noise adaptation for speech recovery with the eigennoise similar to the eigenvoice in voice conversion. The experimental results showed that the proposed approach highly outperformed traditional nonlearning-based approaches.

  7. Very High Resolution Tree Cover Mapping for Continental United States using Deep Convolutional Neural Networks

    Science.gov (United States)

    Ganguly, Sangram; Kalia, Subodh; Li, Shuang; Michaelis, Andrew; Nemani, Ramakrishna R.; Saatchi, Sassan A

    2017-01-01

    Uncertainties in input land cover estimates contribute to a significant bias in modeled above ground biomass (AGB) and carbon estimates from satellite-derived data. The resolution of most currently used passive remote sensing products is not sufficient to capture tree canopy cover of less than ca. 10-20 percent, limiting their utility to estimate canopy cover and AGB for trees outside of forest land. In our study, we created a first of its kind Continental United States (CONUS) tree cover map at a spatial resolution of 1-m for the 2010-2012 epoch using the USDA NAIP imagery to address the present uncertainties in AGB estimates. The process involves different tasks including data acquisition ingestion to pre-processing and running a state-of-art encoder-decoder based deep convolutional neural network (CNN) algorithm for automatically generating a tree non-tree map for almost a quarter million scenes. The entire processing chain including generation of the largest open source existing aerial satellite image training database was performed at the NEX supercomputing and storage facility. We believe the resulting forest cover product will substantially contribute to filling the gaps in ongoing carbon and ecological monitoring research and help quantifying the errors and uncertainties in derived products.

  8. Collaborative classification of hyperspectral and visible images with convolutional neural network

    Science.gov (United States)

    Zhang, Mengmeng; Li, Wei; Du, Qian

    2017-10-01

    Recent advances in remote sensing technology have made multisensor data available for the same area, and it is well-known that remote sensing data processing and analysis often benefit from multisource data fusion. Specifically, low spatial resolution of hyperspectral imagery (HSI) degrades the quality of the subsequent classification task while using visible (VIS) images with high spatial resolution enables high-fidelity spatial analysis. A collaborative classification framework is proposed to fuse HSI and VIS images for finer classification. First, the convolutional neural network model is employed to extract deep spectral features for HSI classification. Second, effective binarized statistical image features are learned as contextual basis vectors for the high-resolution VIS image, followed by a classifier. The proposed approach employs diversified data in a decision fusion, leading to an integration of the rich spectral information, spatial information, and statistical representation information. In particular, the proposed approach eliminates the potential problems of the curse of dimensionality and excessive computation time. The experiments evaluated on two standard data sets demonstrate better classification performance offered by this framework.

  9. Wrist sensor-based tremor severity quantification in Parkinson's disease using convolutional neural network.

    Science.gov (United States)

    Kim, Han Byul; Lee, Woong Woo; Kim, Aryun; Lee, Hong Ji; Park, Hye Young; Jeon, Hyo Seon; Kim, Sang Kyong; Jeon, Beomseok; Park, Kwang S

    2018-04-01

    Tremor is a commonly observed symptom in patients of Parkinson's disease (PD), and accurate measurement of tremor severity is essential in prescribing appropriate treatment to relieve its symptoms. We propose a tremor assessment system based on the use of a convolutional neural network (CNN) to differentiate the severity of symptoms as measured in data collected from a wearable device. Tremor signals were recorded from 92 PD patients using a custom-developed device (SNUMAP) equipped with an accelerometer and gyroscope mounted on a wrist module. Neurologists assessed the tremor symptoms on the Unified Parkinson's Disease Rating Scale (UPDRS) from simultaneously recorded video footages. The measured data were transformed into the frequency domain and used to construct a two-dimensional image for training the network, and the CNN model was trained by convolving tremor signal images with kernels. The proposed CNN architecture was compared to previously studied machine learning algorithms and found to outperform them (accuracy = 0.85, linear weighted kappa = 0.85). More precise monitoring of PD tremor symptoms in daily life could be possible using our proposed method. Copyright © 2018 Elsevier Ltd. All rights reserved.

  10. Deep convolutional neural networks for building extraction from orthoimages and dense image matching point clouds

    Science.gov (United States)

    Maltezos, Evangelos; Doulamis, Nikolaos; Doulamis, Anastasios; Ioannidis, Charalabos

    2017-10-01

    Automatic extraction of buildings from remote sensing data is an attractive research topic, useful for several applications, such as cadastre and urban planning. This is mainly due to the inherent artifacts of the used data and the differences in viewpoint, surrounding environment, and complex shape and size of the buildings. This paper introduces an efficient deep learning framework based on convolutional neural networks (CNNs) toward building extraction from orthoimages. In contrast to conventional deep approaches in which the raw image data are fed as input to the deep neural network, in this paper the height information is exploited as an additional feature being derived from the application of a dense image matching algorithm. As test sites, several complex urban regions of various types of buildings, pixel resolutions and types of data are used, located in Vaihingen in Germany and in Perissa in Greece. Our method is evaluated using the rates of completeness, correctness, and quality and compared with conventional and other "shallow" learning paradigms such as support vector machines. Experimental results indicate that a combination of raw image data with height information, feeding as input to a deep CNN model, provides potentials in building detection in terms of robustness, flexibility, and efficiency.

  11. Very High Resolution Tree Cover Mapping for Continental United States using Deep Convolutional Neural Networks

    Science.gov (United States)

    Ganguly, S.; Kalia, S.; Li, S.; Michaelis, A.; Nemani, R. R.; Saatchi, S.

    2017-12-01

    Uncertainties in input land cover estimates contribute to a significant bias in modeled above gound biomass (AGB) and carbon estimates from satellite-derived data. The resolution of most currently used passive remote sensing products is not sufficient to capture tree canopy cover of less than ca. 10-20 percent, limiting their utility to estimate canopy cover and AGB for trees outside of forest land. In our study, we created a first of its kind Continental United States (CONUS) tree cover map at a spatial resolution of 1-m for the 2010-2012 epoch using the USDA NAIP imagery to address the present uncertainties in AGB estimates. The process involves different tasks including data acquisition/ingestion to pre-processing and running a state-of-art encoder-decoder based deep convolutional neural network (CNN) algorithm for automatically generating a tree/non-tree map for almost a quarter million scenes. The entire processing chain including generation of the largest open source existing aerial/satellite image training database was performed at the NEX supercomputing and storage facility. We believe the resulting forest cover product will substantially contribute to filling the gaps in ongoing carbon and ecological monitoring research and help quantifying the errors and uncertainties in derived products.

  12. Contextual Multi-Scale Region Convolutional 3D Network for Activity Detection

    KAUST Repository

    Bai, Yancheng; Xu, Huijuan; Saenko, Kate; Ghanem, Bernard

    2018-01-01

    Activity detection is a fundamental problem in computer vision. Detecting activities of different temporal scales is particularly challenging. In this paper, we propose the contextual multi-scale region convolutional 3D network (CMS-RC3D) for activity detection. To deal with the inherent temporal scale variability of activity instances, the temporal feature pyramid is used to represent activities of different temporal scales. On each level of the temporal feature pyramid, an activity proposal detector and an activity classifier are learned to detect activities of specific temporal scales. Temporal contextual information is fused into activity classifiers for better recognition. More importantly, the entire model at all levels can be trained end-to-end. Our CMS-RC3D detector can deal with activities at all temporal scale ranges with only a single pass through the backbone network. We test our detector on two public activity detection benchmarks, THUMOS14 and ActivityNet. Extensive experiments show that the proposed CMS-RC3D detector outperforms state-of-the-art methods on THUMOS14 by a substantial margin and achieves comparable results on ActivityNet despite using a shallow feature extractor.

  13. Cerebral microbleed detection in traumatic brain injury patients using 3D convolutional neural networks

    Science.gov (United States)

    Standvoss, K.; Crijns, T.; Goerke, L.; Janssen, D.; Kern, S.; van Niedek, T.; van Vugt, J.; Alfonso Burgos, N.; Gerritse, E. J.; Mol, J.; van de Vooren, D.; Ghafoorian, M.; van den Heuvel, T. L. A.; Manniesing, R.

    2018-02-01

    The number and location of cerebral microbleeds (CMBs) in patients with traumatic brain injury (TBI) is important to determine the severity of trauma and may hold prognostic value for patient outcome. However, manual assessment is subjective and time-consuming due to the resemblance of CMBs to blood vessels, the possible presence of imaging artifacts, and the typical heterogeneity of trauma imaging data. In this work, we present a computer aided detection system based on 3D convolutional neural networks for detecting CMBs in 3D susceptibility weighted images. Network architectures with varying depth were evaluated. Data augmentation techniques were employed to improve the networks' generalization ability and selective sampling was implemented to handle class imbalance. The predictions of the models were clustered using a connected component analysis. The system was trained on ten annotated scans and evaluated on an independent test set of eight scans. Despite this limited data set, the system reached a sensitivity of 0.87 at 16.75 false positives per scan (2.5 false positives per CMB), outperforming related work on CMB detection in TBI patients.

  14. EnzyNet: enzyme classification using 3D convolutional neural networks on spatial representation.

    Science.gov (United States)

    Amidi, Afshine; Amidi, Shervine; Vlachakis, Dimitrios; Megalooikonomou, Vasileios; Paragios, Nikos; Zacharaki, Evangelia I

    2018-01-01

    During the past decade, with the significant progress of computational power as well as ever-rising data availability, deep learning techniques became increasingly popular due to their excellent performance on computer vision problems. The size of the Protein Data Bank (PDB) has increased more than 15-fold since 1999, which enabled the expansion of models that aim at predicting enzymatic function via their amino acid composition. Amino acid sequence, however, is less conserved in nature than protein structure and therefore considered a less reliable predictor of protein function. This paper presents EnzyNet, a novel 3D convolutional neural networks classifier that predicts the Enzyme Commission number of enzymes based only on their voxel-based spatial structure. The spatial distribution of biochemical properties was also examined as complementary information. The two-layer architecture was investigated on a large dataset of 63,558 enzymes from the PDB and achieved an accuracy of 78.4% by exploiting only the binary representation of the protein shape. Code and datasets are available at https://github.com/shervinea/enzynet.

  15. Classification of Land Cover and Land Use Based on Convolutional Neural Networks

    Science.gov (United States)

    Yang, Chun; Rottensteiner, Franz; Heipke, Christian

    2018-04-01

    Land cover describes the physical material of the earth's surface, whereas land use describes the socio-economic function of a piece of land. Land use information is typically collected in geospatial databases. As such databases become outdated quickly, an automatic update process is required. This paper presents a new approach to determine land cover and to classify land use objects based on convolutional neural networks (CNN). The input data are aerial images and derived data such as digital surface models. Firstly, we apply a CNN to determine the land cover for each pixel of the input image. We compare different CNN structures, all of them based on an encoder-decoder structure for obtaining dense class predictions. Secondly, we propose a new CNN-based methodology for the prediction of the land use label of objects from a geospatial database. In this context, we present a strategy for generating image patches of identical size from the input data, which are classified by a CNN. Again, we compare different CNN architectures. Our experiments show that an overall accuracy of up to 85.7 % and 77.4 % can be achieved for land cover and land use, respectively. The classification of land cover has a positive contribution to the classification of the land use classification.

  16. Computer aided detection of ureteral stones in thin slice computed tomography volumes using Convolutional Neural Networks.

    Science.gov (United States)

    Längkvist, Martin; Jendeberg, Johan; Thunberg, Per; Loutfi, Amy; Lidén, Mats

    2018-06-01

    Computed tomography (CT) is the method of choice for diagnosing ureteral stones - kidney stones that obstruct the ureter. The purpose of this study is to develop a computer aided detection (CAD) algorithm for identifying a ureteral stone in thin slice CT volumes. The challenge in CAD for urinary stones lies in the similarity in shape and intensity of stones with non-stone structures and how to efficiently deal with large high-resolution CT volumes. We address these challenges by using a Convolutional Neural Network (CNN) that works directly on the high resolution CT volumes. The method is evaluated on a large data base of 465 clinically acquired high-resolution CT volumes of the urinary tract with labeling of ureteral stones performed by a radiologist. The best model using 2.5D input data and anatomical information achieved a sensitivity of 100% and an average of 2.68 false-positives per patient on a test set of 88 scans. Copyright © 2018 The Authors. Published by Elsevier Ltd.. All rights reserved.

  17. Convolutional deep belief network with feature encoding for classification of neuroblastoma histological images

    Directory of Open Access Journals (Sweden)

    Soheila Gheisari

    2018-01-01

    Full Text Available Background: Neuroblastoma is the most common extracranial solid tumor in children younger than 5 years old. Optimal management of neuroblastic tumors depends on many factors including histopathological classification. The gold standard for classification of neuroblastoma histological images is visual microscopic assessment. In this study, we propose and evaluate a deep learning approach to classify high-resolution digital images of neuroblastoma histology into five different classes determined by the Shimada classification. Subjects and Methods: We apply a combination of convolutional deep belief network (CDBN with feature encoding algorithm that automatically classifies digital images of neuroblastoma histology into five different classes. We design a three-layer CDBN to extract high-level features from neuroblastoma histological images and combine with a feature encoding model to extract features that are highly discriminative in the classification task. The extracted features are classified into five different classes using a support vector machine classifier. Data: We constructed a dataset of 1043 neuroblastoma histological images derived from Aperio scanner from 125 patients representing different classes of neuroblastoma tumors. Results: The weighted average F-measure of 86.01% was obtained from the selected high-level features, outperforming state-of-the-art methods. Conclusion: The proposed computer-aided classification system, which uses the combination of deep architecture and feature encoding to learn high-level features, is highly effective in the classification of neuroblastoma histological images.

  18. A simple method for estimating the length density of convoluted tubular systems.

    Science.gov (United States)

    Ferraz de Carvalho, Cláudio A; de Campos Boldrini, Silvia; Nishimaru, Flávio; Liberti, Edson A

    2008-10-01

    We present a new method for estimating the length density (Lv) of convoluted tubular structures exhibiting an isotropic distribution. Although the traditional equation Lv=2Q/A is used, the parameter Q is obtained by considering the collective perimeters of tubular sections. This measurement is converted to a standard model of the structure, assuming that all cross-sections are approximately circular and have an average perimeter similar to that of actual circular cross-sections observed in the same material. The accuracy of this method was tested in eight experiments using hollow macaroni bent into helical shapes. After measuring the length of the macaroni segments, they were boiled and randomly packed into cylindrical volumes along with an aqueous suspension of gelatin and India ink. The solidified blocks were cut into slices 1.0 cm thick and 33.2 cm2 in area (A). The total perimeter of the macaroni cross-sections so revealed was stereologically estimated using a test system of straight parallel lines. Given Lv and the reference volume, the total length of macaroni in each section could be estimated. Additional corrections were made for the changes induced by boiling, and the off-axis position of the thread used to measure length. No statistical difference was observed between the corrected estimated values and the actual lengths. This technique is useful for estimating the length of capillaries, renal tubules, and seminiferous tubules.

  19. Interpolation between spatial frameworks: an application of process convolution to estimating neighbourhood disease prevalence.

    Science.gov (United States)

    Congdon, Peter

    2014-04-01

    Health data may be collected across one spatial framework (e.g. health provider agencies), but contrasts in health over another spatial framework (neighbourhoods) may be of policy interest. In the UK, population prevalence totals for chronic diseases are provided for populations served by general practitioner practices, but not for neighbourhoods (small areas of circa 1500 people), raising the question whether data for one framework can be used to provide spatially interpolated estimates of disease prevalence for the other. A discrete process convolution is applied to this end and has advantages when there are a relatively large number of area units in one or other framework. Additionally, the interpolation is modified to take account of the observed neighbourhood indicators (e.g. hospitalisation rates) of neighbourhood disease prevalence. These are reflective indicators of neighbourhood prevalence viewed as a latent construct. An illustrative application is to prevalence of psychosis in northeast London, containing 190 general practitioner practices and 562 neighbourhoods, including an assessment of sensitivity to kernel choice (e.g. normal vs exponential). This application illustrates how a zero-inflated Poisson can be used as the likelihood model for a reflective indicator.

  20. Indian Classical Dance Action Identification and Classification with Convolutional Neural Networks

    Directory of Open Access Journals (Sweden)

    P. V. V. Kishore

    2018-01-01

    Full Text Available Extracting and recognizing complex human movements from unconstrained online/offline video sequence is a challenging task in computer vision. This paper proposes the classification of Indian classical dance actions using a powerful artificial intelligence tool: convolutional neural networks (CNN. In this work, human action recognition on Indian classical dance videos is performed on recordings from both offline (controlled recording and online (live performances, YouTube data. The offline data is created with ten different subjects performing 200 familiar dance mudras/poses from different Indian classical dance forms under various background environments. The online dance data is collected from YouTube for ten different subjects. Each dance pose is occupied for 60 frames or images in a video in both the cases. CNN training is performed with 8 different sample sizes, each consisting of multiple sets of subjects. The remaining 2 samples are used for testing the trained CNN. Different CNN architectures were designed and tested with our data to obtain a better accuracy in recognition. We achieved a 93.33% recognition rate compared to other classifier models reported on the same dataset.