检索结果-内蒙古大学图书馆

Scale-regularized filter learning: Calculus of variation meets learning

学校读者我要写书评

暂无评论

arXiv 2017年

作者： Loog, Marco Lauze, François Pattern Recognition Laboratory Delft University of Technology Netherlands Image Section University of Copenhagen Denmark

We start out by demonstrating that an elementary learning task, corresponding to the training of a single linear neuron in a convolutional neural network, can be solved for feature spaces of very high dimensionality. In a second step, acknowledging that such high-dimensional learning tasks typically benefit from some form of regularization and arguing that the problem of scale has not been taken care of in a very satisfactory manner, we come to a combined resolution of both of these shortcomings by proposing a form of scale regularization. Moreover, using variational method, this regularization problem can also be solved rather efficiently and we demonstrate, on an artificial filter learning problem, the capabilities of our basic linear neuron. From a more general standpoint, we see this work as prime example of how learning and variational methods could, or even should work to their mutual benefit. Copyright © 2017, The Authors. All rights reserved.

关键词： Ordinary differential equations

Learning data-adaptive nonparametric kernels

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Liu, Fanghui Huang, Xiaolin Gong, Chen Yang, Jie Li, Li Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University Shanghai200240 China Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information Ministry of Education School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing210094 China Department of Automation Tsinghua University

Kernel methods have been extensively used in a variety of machine learning tasks such as classification, clustering, and dimensionality reduction. For complicated practical tasks, the traditional kernels, e.g., Gaussian kernel and sigmoid kernel, or their combinations are often not sufficiently flexible to fit the data. In this paper, we present a Data-Adaptive Nonparametric Kernel (DANK) learning framework in a data-driven manner. To be specific, in model formulation, we impose an adaptive matrix on the kernel/Gram matrix in an entry-wise strategy. Since we do not specify the formulation of the adaptive matrix, each entry in the adaptive matrix can be directly and flexibly learned from the data. Therefore, the solution space of the learned kernel is largely expanded, which makes our DANK model flexible to capture the data with different local statistical properties. Specifically, the proposed kernel learning framework can be seamlessly embedded to support vector machines (SVM) and support vector regression (SVR), which has the capability of enlarging the margin between classes and reducing the model generalization error. Theoretically, we demonstrate that the objective function of our DANK model embedded in SVM/SVR is gradient-Lipschitz continuous. Thereby, the training process for kernel and parameter learning in SVM/SVR can be efficiently optimized in a unified framework. Further, to address the scalability issue in nonparametric kernel learning framework, we decompose the entire optimization problem in DANK into several smaller easy-to-solve problems, so that our DANK model can be efficiently approximated by this partition. The effectiveness of this approximation is demonstrated by both empirical studies and theoretical guarantees. Experimentally, the proposed DANK model embedded in SVM/SVR achieves encouraging performance on various classification and regression benchmark datasets when compared with other representative kernel learning based algorithms. Copyrig

关键词： Matrix algebra

Correction to: Automatic identification of myopic maculopathy related imaging features in optic disc region via machine learning methods

学校读者我要写书评

暂无评论

Journal of translational medicine 2021年第1期19卷 203页

作者： Yuchen Du Qiuying Chen Ying Fan Jianfeng Zhu Jiangnan He Haidong Zou Dazhen Sun Bowen Xin David Feng Michael Fulham Xiuying Wang Lisheng Wang Xun Xu Department of Automation The Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University (SJTU) 800 Dongchuan RD. Minhang District Shanghai 200240 People's Republic of China. Department of Preventative Ophthalmology Shanghai Eye Diseases Prevention and Treatment Center Shanghai Eye Hospital No. 380 Kangding Road Shanghai 200040 China. Department of Ophthalmology Shanghai Key Laboratory of Ocular Fundus Diseases Shanghai Engineering Center for Visual Science and Photo Medicine Shanghai General Hospital SJTU School of Medicine Shanghai China. National Clinical Research Center for Eye Diseases Shanghai 20080 China. Biomedical and Multimedia Information Technology Research Group School of Computer Science The University of Sydney Sydney NSW 2006 Australia. Department of Molecular Imaging Royal Prince Alfred Hospital and the University of Sydney Sydney Australia. Department of Automation The Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University (SJTU) 800 Dongchuan RD. Minhang District Shanghai 200240 People's Republic of China. lswang@***. Department of Preventative Ophthalmology Shanghai Eye Diseases Prevention and Treatment Center Shanghai Eye Hospital No. 380 Kangding Road Shanghai 200040 China. drxuxun@***. Department of Ophthalmology Shanghai Key Laboratory of Ocular Fundus Diseases Shanghai Engineering Center for Visual Science and Photo Medicine Shanghai General Hospital SJTU School of Medicine Shanghai China. drxuxun@***. National Clinical Research Center for Eye Diseases Shanghai 20080 China. drxuxun@***.

An amendment to this paper has been published and can be accessed via the original article.

关键词：

Fast signal recovery from saturated measurements by linear loss and nonconvex penalties

学校读者我要写书评

暂无评论

arXiv 2018年

作者： He, Fan Huang, Xiaolin Liu, Yipeng Yan, Ming Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University The MOE Key Laboratory of System Control and Information Processing Shanghai200240 China School of Information and Communication Engineering University of Electronic Science and Technology of China Chengdu611731 China The Department of Computational Mathematics Science and Engineering Michigan State University MI United States

Sign information is the key to overcoming the inevitable saturation error in compressive sensing systems, which causes information loss and results in bias. For sparse signal recovery from saturation, we propose to use a linear loss to improve the effectiveness from existing methods that utilize hard constraints/hinge loss for sign consistency. Due to the use of linear loss, an analytical solution in the update progress is obtained, and some nonconvex penalties are applicable, e.g., the minimax concave penalty, the 0 norm, and the sorted 1 norm. Theoretical analysis reveals that the estimation error can still be bounded. Generally, with linear loss and nonconvex penalties, the recovery performance is significantly improved, and the computational time is largely saved, which is verified by the numerical experiments. Copyright © 2018, The Authors. All rights reserved.

关键词： Compressed sensing

Generalization Properties of hyper-RKHS and its Applications

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Liu, Fanghui Shi, Lei Huang, Xiaolin Yang, Jie Suykens, Johan A.K. Department of Electrical Engineering ESAT-STADIUS KU Leuven Kasteelpark Arenberg 10 LeuvenB-3001 Belgium Shanghai Key Laboratory for Contemporary Applied Mathematics School of Mathematical Sciences Fudan University Shanghai200433 China Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University Institute of Medical Robotics Shanghai Jiao Tong University Shanghai200240 China

This paper generalizes regularized regression problems in a hyper-reproducing kernel Hilbert space (hyper-RKHS), illustrates its utility for kernel learning and out-of-sample extensions, and proves asymptotic convergence results for the introduced regression models in an approximation theory view. Algorithmically, we consider two regularized regression models with bivariate forms in this space, including kernel ridge regression (KRR) and support vector regression (SVR) endowed with hyper-RKHS, and further combine divide-and-conquer with Nyström approximation for scalability in large sample cases. This framework is general: the underlying kernel is learned from a broad class, and can be positive definite or not, which adapts to various requirements in kernel learning. Theoretically, we study the convergence behavior of regularized regression algorithms in hyper-RKHS and derive the learning rates, which goes beyond the classical analysis on RKHS due to the non-trivial independence of pairwise samples and the characterisation of hyper-RKHS. Experimentally, results on several benchmarks suggest that the employed framework is able to learn a general kernel function form an arbitrary similarity matrix, and thus achieves a satisfactory performance on classification tasks. Copyright © 2018, The Authors. All rights reserved.

关键词： Approximation theory

3D RoI-aware U-net for accurate and efficient colorectal tumor segmentation

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Huang, Yi-Jie Dou, Qi Wang, Zi-Xian Liu, Li-Zhi Jin, Ying Li, Chao-Feng Wang, Lisheng Chen, Hao Xu, Rui-Hua Institute of Image Processing and Pattern Recognition Department of Automation Shanghai Jiao Tong University China Imsight Medical Technology Co. Ltd. China Department of Computer Science and Engineering The Chinese University of Hong Kong Hong Kong Sun Yat-sen University Cancer Center State Key Laboratory of Oncology in South China Collaborative Innovation Center for Cancer Medicine Guangzhou China

Segmentation of colorectal cancerous regions from 3D Magnetic Resonance (MR) images is a crucial procedure for radiotherapy which conventionally requires accurate delineation of tumour boundaries at an expense of labor, time and reproducibility. While deep learning based methods serve good baselines in 3D image segmentation tasks, small applicable patch size limits effective receptive field and degrades segmentation performance. In addition, Regions of interest (RoIs) localization from large whole volume 3D images serves as a preceding operation that brings about multiple benefits in terms of speed, target completeness, reduction of false positives. Distinct from sliding window or non-joint localization-segmentation based models, we propose a novel multitask framework referred to as 3D RoI-aware U-Net (3D RU-Net), for RoI localization and in-region segmentation where the two tasks share one backbone encoder network. With the region proposals from the encoder, we crop multi-level RoI in-region features from the encoder to form a GPU memory-efficient decoder for detail-preserving segmentation and therefore enlarged applicable volume size and effective receptive field. To effectively train the model, we designed a Dice formulated loss function for the global-to-local multi-task learning procedure. Based on the efficiency gains demonstrated by the proposed method, we went on to ensemble models with different receptive fields to achieve even higher performance costing minor extra computational expensiveness. Extensive experiments were subsequently conducted on 64 cancerous cases with a four-fold cross-validation, and the results showed significant superiority in terms of accuracy and efficiency over conventional state-of-the art frameworks. In conclusion, the proposed method has a huge potential for extension to other 3D object segmentation tasks from medical images due to its inherent generalizability. The code for the proposed method is publicly available. Copyright ©

关键词： Signal encoding

Edgy salient local binary patterns in inter-plane relationship for image retrieval in Diabetic Retinopathy

学校读者我要写书评

暂无评论

Procedia Computer Science 2017年 115卷 440-447页

作者： Gajanan M. Galshetwar Laxman M. Waghmare Anil B. Gonde Subrahmanyam Murala Center of Excellence in Signal and Image Processing (COESIP) Department of ECE SGGSIET Nanded Maharashtra 431606 India Computer Vision and Pattern Recognition Laboratory Department of Electrical Engineering IIT Ropar Rupnagar 140001 India

In this paper, a novel approach for content based image retrieval (CBIR) in diabetic retinopathy (DR) is proposed. The concept of salient point selection and inter-plane relationship technique is used. Salient points are selected from edgy image and later using inter-planer relationship, Local Binary patterns (LBPs) are calculated using the salient point as a center pixel. Our approach enhanced the results as we used color features in combination with LBP features. Experimentation is carried out on MESSIDOR database of 1200 retinal images, proposed approach has average precision of 57.82% as compared to the earlier approach whose average precision is 53.70%.

关键词： Content-Based image retrieval (CBIR) Diabetic Retinopathy (DR) Edgy Salient points Local Binary patterns (LBPs)

Label stability in multiple instance learning

学校读者我要写书评

暂无评论

arXiv 2017年

作者： Cheplygina, Veronika Sørensen, Lauge Tax, David M.J. de Bruijne, Marleen Loog, Marco Pattern Recognition Laboratory Delft University of Technology Netherlands Image Section University of Copenhagen Copenhagen Denmark Biomedical Imaging Group Rotterdam Erasmus MC Rotterdam Netherlands

We address the problem of instance label stability in multiple instance learning (MIL) classifiers. These classifiers are trained only on globally annotated images (bags), but often can provide fine-grained annotations for image pixels or patches (instances). This is interesting for computer aided diagnosis (CAD) and other medical image analysis tasks for which only a coarse labeling is provided. Unfortunately, the instance labels may be unstable. This means that a slight change in training data could potentially lead to abnormalities being detected in different parts of the image, which is undesirable from a CAD point of view. Despite MIL gaining popularity in the CAD literature, this issue has not yet been addressed. We investigate the stability of instance labels provided by several MIL classifiers on 5 different datasets, of which 3 are medical image datasets (breast histopathology, diabetic retinopathy and computed tomography lung images). We propose an unsupervised measure to evaluate instance stability, and demonstrate that a performance-stability trade-off can be made when comparing MIL classifiers. Copyright © 2017, The Authors. All rights reserved.

关键词： Computer aided diagnosis

Target contrastive pessimistic risk for robust domain adaptation

学校读者我要写书评

暂无评论

arXiv 2017年

作者： Kouw, Wouter M. Loog, Marco Pattern Recognition Laboratory Delft University of Technology Mekelweg 4 Delft2628 CD Netherlands Image Group University of Copenhagen Universitetsparken 5 CopenhagenDK-2100 Denmark

In domain adaptation, classifiers with information from a source domain adapt to generalize to a target domain. However, an adaptive classifier can perform worse than a non-adaptive classifier due to invalid assumptions, increased sensitivity to estimation errors or model misspecification. Our goal is to develop a domain-adaptive classifier that is robust in the sense that it does not rely on restrictive assumptions on how the source and target domains relate to each other and that it does not perform worse than the non-adaptive classifier. We formulate a conservative parameter estimator that only deviates from the source classifier when a lower risk is guaranteed for all possible labellings of the given target samples. We derive the classical least-squares and discriminant analysis cases and show that these perform on par with state-of-the-art domain adaptive classifiers in sample selection bias settings, while outperforming them in more general domain adaptation settings. Copyright © 2017, The Authors. All rights reserved.

关键词： Discriminant analysis