检索结果-内蒙古大学图书馆

End-to-end, single-stream temporal action detection in untrimmed videos 28

学校读者我要写书评

暂无评论

End-to-end, single-stream temporal action detection in untri...

28th British Machine Vision Conference, BMVC 2017

作者： Buch, Shyamal Escorcia, Victor Ghanem, Bernard Fei-Fei, Li Niebles, Juan Carlos Stanford Vision and Learning Lab. Dept. of Computer Science Stanford University United States Image and Video Understanding Lab. Visual Computing Center KAUST Saudi Arabia

ISBN: (纸本)190172560X

In this work, we present a new intuitive, end-to-end approach for temporal action detection in untrimmed videos. We introduce our new architecture for Single-Stream Temporal Action Detection (SS-TAD), which effectively integrates joint action detection with its semantic sub-tasks in a single unifying end-to-end framework. We develop a method for training our deep recurrent architecture based on enforcing semantic constraints on intermediate modules that are gradually relaxed as learning progresses. We find that such a dynamic learning scheme enables SS-TAD to achieve higher overall detection performance, with fewer training epochs. By design, our single-pass network is very efficient and can operate at 701 frames per second, while simultaneously outperforming the state-of-the-art methods for temporal action detection on THUMOS’14. © 2017. The copyright of this document resides with its authors.

关键词： Semantics

Deep learning on image denoising: An overview

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Tian, Chunwei Fei, Lunke Zheng, Wenxian Xu, Yong Zuo, Wangmeng Lin, Chia-Wen Bio-Computing Research Center Harbin Institute of Technology Shenzhen Shenzhen Guangdong518055 China Shenzhen Key Laboratory of Visual Object Detection and Recognition Shenzhen Guangdong518055 China School of Computers Guangdong University of Technology Guangzhou Guangdong510006 China Tsinghua Shenzhen International Graduate School Shenzhen Guangdong518055 China Peng Cheng Laboratory Shenzhen Guangdong518055 China School of Computer Science and Technology Harbin Institute of Technology Harbin Heilongjiang150001 China Department of Electrical Engineering Institute of Communications Engineering National Tsing Hua University Hsinchu Taiwan

Deep learning techniques have obtained much attention in image denoising. However, deep learning methods of different types deal with the noise have enormous differences. Specifically, discriminative learning based on deep learning can well address the Gaussian noise. Optimization model methods based on deep learning have good effect on estimating of the real noise. So far, there are little related researches to summarize different deep learning techniques for image denoising. In this paper, we make such a comparative study of different deep techniques in image denoising. We first classify the (1) deep convolutional neural networks (CNNs) for additive white noisy images, (2) deep CNNs for real noisy images, (3) deep CNNs for blind denoising and (4) deep CNNs for hybrid noisy images, which is the combination of noisy, blurred and low-resolution images. Then, we analyze the motivations and principles of deep learning methods of different types. Next, we compare and verify the state-of-the-art methods on public denoising datasets in terms of quantitative and qualitative analysis. Finally, we point out some potential challenges and directions of future research. Copyright © 2019, The Authors. All rights reserved.

关键词： Deep neural networks

A two-stage subspace trust region approach for deep neural network training

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Dudar, Viacheslav Chierchia, Giovanni Chouzenoux, Emilie Pesquet, Jean-Christophe Semenov, Vladimir Taras Shevchenko National University of Kyiv Faculty of Computer Science and Cybernetics Ukraine Université Paris Est LIGM UMR 8049 CNRS ENPC ESIEE Paris UPEM Noisy-le-GrandF-93162 France Center for Visual Computing CentraleSupelec University Paris-Saclay Chatenay-Malabry France

—In this paper, we develop a novel second-order method for training feed-forward neural nets. At each iteration, we construct a quadratic approximation to the cost function in a low-dimensional subspace. We minimize this approximation inside a trust region through a two-stage procedure: first inside the embedded positive curvature subspace, followed by a gradient descent step. This approach leads to a fast objective function decay, prevents convergence to saddle points, and alleviates the need for manually tuning parameters. We show the good performance of the proposed algorithm on benchmark datasets. Copyright © 2018, The Authors. All rights reserved.

关键词： Gradient methods

A probabilistic incremental proximal gradient method

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Akyildiz, Ömer Deniz Chouzenoux, Émilie Elvira, Víctor Míguez, Joaquín Dept. of Computer Science Dept. of Statistics University of Warwick Alan Turing Institute London United Kingdom Center for Visual Computing INRIA Saclay CentraleSupélec Gif-sur-Yvette France Villeneuve d’Ascq France Dept. of Signal Theory and Communications Universidad Carlos III de Madrid Leganés28912 Spain

In this paper, we propose a probabilistic optimization method, named probabilistic incremental proximal gradient (PIPG) method, by developing a probabilistic interpretation of the incremental proximal gradient algorithm. We explicitly model the update rules of the incremental proximal gradient method and develop a systematic approach to propagate the uncertainty of the solution estimate over iterations. The PIPG algorithm takes the form of Bayesian filtering updates for a state-space model constructed by using the cost function. Our framework makes it possible to utilize well-known exact or approximate Bayesian filters, such as Kalman or extended Kalman filters, to solve large-scale regularized optimization problems. Copyright © 2018, The Authors. All rights reserved.

关键词： Gradient methods

A two-stage subspace trust region approach for deep neural network training

学校读者我要写书评

暂无评论

A two-stage subspace trust region approach for deep neural n...

European Signal Processing Conference (EUSIPCO)

作者： Viacheslav Dudar Giovanni Chierchia Emilie Chouzenoux Jean-Christophe Pesquet Vladimir Semenov Faculty of Computer Science and Cybernetics Taras Shevchenko National University of Kyiv Ukraine Universite Paris Est UPEM Noisy-le-Grand France Center for Visual Computing University Paris-Saclay Chatenay-Malabry France Center for Visual Computing CentraleSupelec University Paris-Saclay Chatenay-Malabry France

In this paper, we develop a novel second-order method for training feed-forward neural nets. At each iteration, we construct a quadratic approximation to the cost function in a low-dimensional subspace. We minimize this approximation inside a trust region through a two-stage procedure: first inside the embedded positive curvature subspace, followed by a gradient descent step. This approach leads to a fast objective function decay, prevents convergence to saddle points, and alleviates the need for manually tuning parameters. We show the good performance of the proposed algorithm on benchmark datasets.

关键词： Training Neural networks Signal processing algorithms Optimization Eigenvalues and eigenfunctions Europe Signal processing

Automatic image cropping for visual aesthetic enhancement using deep neural networks and cascaded regression

学校读者我要写书评

暂无评论

arXiv 2017年

作者： Guo, Guanjun Wang, Hanzi Shen, Chunhua Yan, Yan Liao, Hong-Yuan Mark Fujian Key Laboratory of Sensing and Computing for Smart City School of Information Science and Engineering Xiamen University Xiamen Fujian361005 China Australian Center for Visual Technologies School of Computer Science University of Adelaide SA5005 Australia Institute of Information Science Academia Sinica Taipei115 Taiwan

Despite recent progress, computational visual aesthetic is still challenging. Image cropping, which refers to the removal of unwanted scene areas, is an important step to improve the aesthetic quality of an image. However, it is challenging to evaluate whether cropping leads to aesthetically pleasing results because the assessment is typically subjective. In this paper, we propose a novel cascaded cropping regression (CCR) method to perform image cropping by learning the knowledge from professional photographers. The proposed CCR method improves the convergence speed of the cascaded method, which directly uses random-ferns regressors. In addition, a two-step learning strategy is proposed and used in the CCR method to address the problem of lacking labelled cropping data. Specifically, a deep convolutional neural network (CNN) classifier is first trained on large-scale visual aesthetic datasets. The deep CNN model is then designed to extract features from several image cropping datasets, upon which the cropping bounding boxes are predicted by the proposed CCR method. Experimental results on public image cropping datasets demonstrate that the proposed method significantly outperforms several state-of-the-art image cropping methods. Copyright © 2017, The Authors. All rights reserved.

关键词： Regression analysis

A novel dual ascent algorithm for solving the min-cost flow problem 18

学校读者我要写书评

暂无评论

A novel dual ascent algorithm for solving the min-cost flow ...

18th Workshop on Algorithm Engineering and Experiments 2016, ALENEX 2016

作者： Becker, Ruben Fickert, Maximilian Karrenbauer, Andreas Max Planck Institute for Informatics Saarbriicken Germany Max Planck Center for Visual Computing and Communication Saarbriicken Germany Saarbriicken Graduate School of Computer Science Saarbriicken Germany

ISBN: (纸本)9781510819689

We present a novel alg orithm for the min-cost flow problem that is competitive with recent third-party implementations of well-known algorithms for this problem and even outperforms them on certain realistic instances. We formally prove correctness of our algorithm and show that the worst-case running time is in 0(||β||i(m + ηlogη)) where b is the vector of demands. Combined with standard scaling techniques, this pseudo-polynomial bound can be made polynomial in a straightforward way. Furthermore, we evaluate our approach experimentally. Our empirical findings indeed suggest that the running time does not significantly depend on the costs and that a linear dependence on ||β||i is overly pessimistic.

关键词： Cost accounting

Predicting breast tumor proliferation from whole-slide images: The TUPAC16 challenge

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Veta, Mitko Heng, Yujing J. Stathonikos, Nikolas Bejnordi, Babak Ehteshami Beca, Francisco Wollmann, Thomas Rohr, Karl Shah, Manan A. Wang, Dayong Rousson, Mikael Hedlund, Martin Tellez, David Ciompi, Francesco Zerhouni, Erwan Lanyi, David Viana, Matheus Kovalev, Vassili Liauchuk, Vitali Phoulady, Hady Ahmady Qaiser, Talha Graham, Simon Rajpoot, Nasir Sjöblom, Erik Molin, Jesper Paeng, Kyunghyun Hwang, Sangheum Park, Sunggyun Jia, Zhipeng Chang, Eric I-Chao Xu, Yan Beck, Andrew H. Van Diest, Paul J. Pluim, Josien P.W. Medical Image Analysis Group Eindhoven University of Technology Eindhoven Netherlands Department of Pathology Beth Israel Deaconess Medical Center Harvard Medical School BostonMA United States Department of Pathology University Medical Center Utrecht Utrecht Netherlands Diagnostic Image Analysis Group Radboud University Medical Center Nijmegen Netherlands Department of Pathology Stanford University School of Medicine Biomedical Computer Vision Group University of Heidelberg BIOQUANT IPMB DKFZ Heidelberg Germany Harker School San Jose United States ContextVision AB Linköping Sweden Foundations of Cognitive Computing IBM Research Zürich Rüschlikon Switzerland Visual Analytics and Insights IBM Research Brazil Saõ Paulo Brazil Biomedical Image Analysis Department United Institute of Informatics Minsk Belarus Department of Computer Science and Engineering University of South Florida TampaFL United States Department of Computer Science University of Warwick Warwick United Kingdom Research Sectra Linköping Sweden Lunit Inc. Seoul Korea Republic of Institute for Interdisciplinary Information Sciences Tsinghua University Beijing China Microsoft Research Beijing China Biology and Medicine Engineering Beihang University Beijing China IBM Research Brazil

Tumor proliferation is an important biomarker indicative of the prognosis of breast cancer patients. Assessment of tumor proliferation in a clinical setting is a highly subjective and labor-intensive task. Previous efforts to automate tumor proliferation assessment by image analysis only focused on mitosis detection in predefined tumor regions. However, in a realworld scenario, automatic mitosis detection should be performed in whole-slide images (WSIs) and an automatic method should be able to produce a tumor proliferation score given a WSI as input. To address this, we organized the TUmor Proliferation Assessment Challenge 2016 (TUPAC16) on prediction of tumor proliferation scores from WSIs. The challenge dataset consisted of 500 training and 321 testing breast cancer histopathology WSIs. In order to ensure fair and independent evaluation, only the ground truth for the training dataset was provided to the challenge participants. The first task of the challenge was to predict mitotic scores, i.e., to reproduce the manual method of assessing tumor proliferation by a pathologist. The second task was to predict the gene expression based PAM50 proliferation scores from the WSI. The best performing automatic method for the first task achieved a quadratic-weighted Cohen's kappa score of κ = 0.567, 95% CI [0.464, 0.671] between the predicted scores and the ground truth. For the second task, the predictions of the top method had a Spearman's correlation coefficient of r = 0.617, 95% CI [0.581 0.651] with the ground truth. This was the first comparison study that investigated tumor proliferation assessment from WSIs. The achieved results are promising given the difficulty of the tasks and weakly-labeled nature of the ground truth. However, further research is needed to improve the practical utility of image analysis methods for this task. Copyright © 2018, The Authors. All rights reserved.

关键词： Diseases

Symbolic-numeric integration of the dynamical cosserat equations

学校读者我要写书评

暂无评论

arXiv 2017年

作者： Lyakhov, Dmitry A. Gerdt, Vladimir P. Weber, Andreas G. Michels, Dominik L. Visual Computing Center King Abdullah University of Science and Technology Al Khawarizmi Building Thuwal23955-6900 Saudi Arabia Department of Computer Science Stanford University 353 Serra Mall StanfordCA94305 United States Laboratory of Information Technologies Joint Institute for Nuclear Research 6 Joliot-Curie St Dubna141980 Russia Peoples' Friendship University of Russia 6 Miklukho-Maklaya St Moscow117198 Russia Institute of Computer Science II University of Bonn Friedrich-Ebert-Allee 144 Bonn53113 Germany

We devise a symbolic-numeric approach to the integration of the dynamical part of the Cosserat equations, a system of nonlinear partial differential equations describing the mechanical behavior of slender structures, like fibers and rods. This is based on our previous results on the construction of a closed form general solution to the kinematic part of the Cosserat system. Our approach combines methods of numerical exponential integration and symbolic integration of the intermediate system of nonlinear ordinary differential equations describing the dynamics of one of the arbitrary vector-functions in the general solution of the kinematic part in terms of the module of the twist vector-function. We present an experimental comparison with the well-established generalized α-method illustrating the computational efficiency of our approach for problems in structural mechanics. Copyright © 2017, The Authors. All rights reserved.

关键词： Integration