检索结果-内蒙古大学图书馆

arXiv 2018年

作者： He, Lei Wang, Guanghui Hu, Zhanyi University of Chinese Academy of Sciences Beijing100049 China National Laboratory of Pattern Recognition Institute of Automation Chinese Academy of Sciences Beijing100190 China CAS Center for Excellence in Brain Science and Intelligence Technology Chinese Academy of Sciences Beijing100190 China Department of Electrical Engineering and Computer Science University of Kansas LawrenceKS66045 United States

Learning depth from a single image, as an important issue in scene understanding, has attracted a lot of attention in the past decade. The accuracy of the depth estimation has been improved from conditional Markov random fields, non-parametric methods, to deep convolutional neural networks most recently. However, there exist inherent ambiguities in recovering 3D from a single 2D image. In this paper, we first prove the ambiguity between the focal length and monocular depth learning, and verify the result using experiments, showing that the focal length has a great influence on accurate depth recovery. In order to learn monocular depth by embedding the focal length, we propose a method to generate synthetic varying-focal-length dataset from fixed-focal-length datasets, and a simple and effective method is implemented to fill the holes in the newly generated images. For the sake of accurate depth recovery, we propose a novel deep neural network to infer depth through effectively fusing the middle-level information on the fixed-focal-length dataset, which outperforms the state-of-the-art methods built on pre-trained VGG. Furthermore, the newly generated varying-focal-length dataset is taken as input to the proposed network in both learning and inference phases. Extensive experiments on the fixed- and varying-focal-length datasets demonstrate that the learned monocular depth with embedded focal length is significantly improved compared to that without embedding the focal length information. Copyright © 2018, The Authors. All rights reserved.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

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

arXiv

引用

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

来源：评论

学校读者我要写书评

暂无评论

Deep crisp boundaries: From boundaries to higher-level tasks

arXiv

引用

arXiv 2018年

作者： Wang, Yupei Zhao, Xin Li, Yin Huang, Kaiqi Center for Research on Intelligent System and Engineering Institute of Automation Chinese Academy of Sciences Beijing100190 China University of Chinese Academy of Sciences Beijing100049 China Department of Biostatistics and Medical Informatics Department of Computer Sciences Univeristy of Wisconsin-Madison Center for Research on Intelligent System and Engineering National Laboratory of Pattern Recognition Institute of Automation Chinese Academy of Sciences Beijing100190 China University of Chinese Academy of Sciences Beijing100049 China CAS Center for Excellence in Brain Science and Intelligence Technology 100190

Edge detection has made significant progress with the help of deep Convolutional Networks (ConvNet). These ConvNet based edge detectors have approached human level performance on standard benchmarks. We provide a systematical study of these detectors' outputs. We show that the detection results did not accurately localize edge pixels, which can be adversarial for tasks that require crisp edge inputs. As a remedy, we propose a novel refinement architecture to address the challenging problem of learning a crisp edge detector using ConvNet. Our method leverages a top-down backward refinement pathway, and progressively increases the resolution of feature maps to generate crisp edges. Our results achieve superior performance, surpassing human accuracy when using standard criteria on BSDS500, and largely outperforming state-of-the-art methods when using more strict criteria. More importantly, we demonstrate the benefit of crisp edge maps for several important applications in computer vision, including optical flow estimation, object proposal generation and semantic segmentation. Copyright © 2018, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Dynamic vortical flow profile quantitatively characterizes mitral valvular-left ventricular hemodynamic coupling: In vivo analysis by dynamic 3D enstrophy mapping from 4D Flow MRI

arXiv

引用

arXiv 2018年

作者： Elbaz, Mohammed S.M. Le, Trung Bao Garg, Pankaj Roest, Arno A.W. Lelieveldt, Boudewijn P.F. Plein, Sven Westenberg, Jos J.M. Sotiropoulos, Fotis van der Geest, Rob J. Department of Radiology Northwestern University ChicagoIL United States Department of Radiology Leiden University Medical Center Leiden Netherlands Department of Civil Engineering Stony Brook University Stony BrookNY United States University of Leeds Leeds United Kingdom Department of Pediatrics Division of Pediatric Cardiology Leiden University Medical Center Leiden Netherlands Pattern Recognition and Bioinformatics Group Faculty of Electrical Engineering Mathematics and Computer Science Delft University of Technology Delft Netherlands Computational Fluids Laboratory North Dakota State University United States

Vortical blood flow in the human left ventricular (LV) inflow initiates from the mitral valve (MV) and evolves within the LV during diastolic E-filling. Hence, vortical flow links MV and LV hemodynamics. This study sought to elucidate and quantitatively characterize the in vivo 3D dynamics of LV vortical flow over E-filling and relation to MV-LV hemodynamic coupling using 4D Flow MRI flow field. 34 healthy volunteers and 5 example patients underwent 4D Flow MRI. Vortical blood flow evolution was mapped in the LV over E-wave using enstrophy density. A new dimensionless profile PMV-LV was derived as a function of both MV vortex formation time (VFT) and LV volumetric enstrophy density. Results reveal that 3D vortical flow evolution in the healthy LV follows a bi-phasic behavior with a vortical growth phase followed by a vortical decay phase. In healthy LVs studied, the PMV-LV profile showed that the vortical growth and decay phases are characterized by a vortical growth time Tgrowth = 1.23 ± 0.25, growth rate α = 0.80 ± 0.17, decay time Tdecay = 0.96 ± 0.39 and decay rate β = -1.02 ± 0.49. Distinctly altered parameters were found in the pilot patients studied. The derived PMV-LV profile quantitatively characterizes MV-LV hemodynamic coupling by vortical flow dynamics. Results herein unravel new insights into cardiac physiology and could enable a novel standardized methodology to study MV-LV hemodynamic coupling and association to cardiac function in health and disease. Copyright © 2018, The Authors. All rights reserved.

关键词： Heart

来源：评论

学校读者我要写书评

暂无评论

Face alignment across large poses: A 3D solution

Face alignment across large poses: A 3D solution

引用

2016 IEEE Conference on computer Vision and pattern recognition, CVPR 2016

作者： Zhu, Xiangyu Lei, Zhen Liu, Xiaoming Shi, Hailin Li, Stan Z. Center for Biometrics and Security Research National Laboratory of Pattern Recognition Institute of Automation Chinese Academy of Sciences China Department of Computer Science and Engineering Michigan State University United States

ISBN: (纸本)9781467388511

Face alignment, which fits a face model to an image and extracts the semantic meanings of facial pixels, has been an important topic in CV community. However, most algorithms are designed for faces in small to medium poses (below 45°), lacking the ability to align faces in large poses up to 90°. The challenges are three-fold: Firstly, the commonly used landmark-based face model assumes that all the landmarks are visible and is therefore not suitable for profile views. Secondly, the face appearance varies more dramatically across large poses, ranging from frontal view to profile view. Thirdly, labelling landmarks in large poses is extremely challenging since the invisible landmarks have to be guessed. In this paper, we propose a solution to the three problems in an new alignment framework, called 3D Dense Face Alignment (3DDFA), in which a dense 3D face model is fitted to the image via convolutional neutral network (CNN). We also propose a method to synthesize large-scale training samples in profile views to solve the third problem of data labelling. Experiments on the challenging AFLW database show that our approach achieves significant improvements over state-of-the-art methods.

关键词： Alignment

来源：评论

学校读者我要写书评

暂无评论

A Score Recommendation System Towards Automating Assessment In Professional Courses

A Score Recommendation System Towards Automating Assessment ...

引用

International Workshop on Technology for Education, T4E

作者： Anirudh Kashi Sachin Shastri Akshay R. Deshpande Jawahar Doreswamy Gowri Srinivasa PES Center for Pattern Recognition Dept. of Computer Science and Engineering PES Inst. of Technology Bengaluru India Dept. of Information Science and Engineering PES Inst. of Technology Bengaluru India PES University Bengaluru India Dept. of Computer Science and Engineering PESIT Bangalore South Campus Bengaluru India

ISBN: (纸本)9781509061167

We present a preliminary version of an automated evaluation system that provides recommended scores for descriptive answers. The system is designed in the context of professional courses, such as from Applied sciences, for questions that have an answer-key. The recommendation of a score is based on an analysis of the student's answer against an answer-key that comprises key phrases of the solution and alternatives, if any. We rely on five algorithms from literature on natural language processing to assess various aspects of a student's answer, such as the quantity and extent of match between keywords in a student's answer and the answer-key, while ensuring merely verbatim comparisons are not performed. The similarity scores are then combined using machine learning to output a recommended score for the answer. The system learns weights for the output of each of the five methods to map these to a score based on training data, comprising questions with answer-keys, student answers from actual exams and the scores provided by course instructors/ graders. It is noteworthy that the methods presented in this paper do not rely on any kind of a corpus or prior-knowledge that is subject-specific. Even though the accuracies are modest and improvements ought to be made to accommodate the kind of variety observed in students' answers, the avenue is a promising one towards enhancing learning assessment.

关键词： Mathematical model Logistics Semantics Plagiarism Dictionaries Matrix decomposition Frequency measurement

来源：评论

学校读者我要写书评

暂无评论

Multimodal Fusion of Brain Imaging Data: A Key to Finding the Missing Link(s) in Complex Mental Illness

引用

Biological Psychiatry: Cognitive Neuroscience and Neuroimaging 2016年第3期1卷 230-244页

作者： Calhoun, Vince D. Sui, Jing Mind Research Network Lovelace Biomedical and Environmental Research Institute Albuquerque NM United States Department of Electrical and Computer Engineering University of New Mexico Albuquerque NM United States Brainnetome Center National Laboratory of Pattern Recognition Beijing China Chinese Academy of Sciences Center for Excellence in Brain Science Institute of Automation Beijing China

It is becoming increasingly clear that combining multimodal brain imaging data provides more information for individual subjects by exploiting the rich multimodal information that exists. However, the number of studies that do true multimodal fusion (i.e., capitalizing on joint information among modalities) is still remarkably small given the known benefits. In part, this is because multimodal studies require broader expertise in collecting, analyzing, and interpreting the results than do unimodal studies. In this article, we start by introducing the basic reasons why multimodal data fusion is important and what it can do and, importantly, how it can help us avoid wrong conclusions and help compensate for imperfect brain imaging studies. We also discuss the challenges that need to be confronted for such approaches to be more widely applied by the community. We then provide a review of the diverse studies that have used multimodal data fusion (primarily focused on psychosis) as well as provide an introduction to some of the existing analytic approaches. Finally, we discuss some up-and-coming approaches to multimodal fusion including deep learning and multimodal classification that show considerable promise. Our conclusion is that multimodal data fusion is rapidly growing, but it is still underutilized. The complexity of the human brain coupled with the incomplete measurement provided by existing imaging technology makes multimodal fusion essential to mitigate misdirection and hopefully provide a key to finding the missing link(s) in complex mental illness. © 2016 Society of Biological Psychiatry.

关键词： Brain function Connectivity Data fusion Independent component analysis Psychosis Schizophrenia

来源：评论

学校读者我要写书评

暂无评论

Constructing multi-modality and multi-classifier radiomics predictive models through reliable classifier fusion

arXiv

引用

arXiv 2017年

作者： Zhou, Zhiguo Zhou, Zhi-Jie Hao, Hongxia Li, Shulong Chen, Xi Zhang, You Folkert, Michael Wang, Jing Department of Radiation Oncology UT Southwestern Medical Center DallasTX United States High-Tech Institute of Xi'an Xi'an China Department of Information and Control Engineering Xi'an University of Technology Xi'an China School of Computer Science and Technology Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education Xidian University Xi'an China School of Biomedical Engineering Southern Medical Unversity Guangzhou China Institute of Image Processing and Pattern Recognition Xi'an Jiaotong Unversity Xi'an China

Radiomics aims to extract and analyze large numbers of quantitative features from medical images and is highly promising in staging, diagnosing, and predicting outcomes of cancer treatments. Nevertheless, several challenges need to be addressed to construct an optimal radiomics predictive model. First, the predictive performance of the model may be reduced when features extracted from an individual imaging modality (e.g. PET, CT, MRI) are blindly combined into a single predictive model. Second, because many different types of classifiers are available to construct a predictive model, selecting an "optimal" classifier for a particular application is still challenging. In this work, we developed multi-modality and multi-classifier radiomics predictive models that address the aforementioned issues in currently available models. Specifically, a new reliable classifier fusion strategy was proposed to optimally combine output from different modalities and classifiers. In this strategy, modality-specific classifiers were first trained, and an analytic evidential reasoning (ER) rule was developed to fuse the output score from each modality to construct an optimal predictive model. One public datasets and two clinical case studies were performed to validate model performance. The experimental results indicated that the proposed ER rule based radiomics models outperformed the traditional models that rely on a single classifier or simply use combined features from different modalities. Copyright © 2017, The Authors. All rights reserved.

关键词： Diagnosis

来源：评论

学校读者我要写书评

暂无评论

Face Alignment Across Large Poses: A 3D Solution

Face Alignment Across Large Poses: A 3D Solution

引用

IEEE Conference on computer Vision and pattern recognition

作者： Xiangyu Zhu Zhen Lei Xiaoming Liu Hailin Shi Stan Z. Li Center for Biometrics and Security Research & National Laboratory of Pattern Recognition Institute of Automation Chinese Academy of Sciences Department of Computer Science and Engineering Michigan State University

ISBN: (纸本)9781467388528

关键词： face modeling Face Posing Face landmarks Imagery (Psychotherapy) images Alignment

来源：评论

学校读者我要写书评

暂无评论

A Survey of Blue-Noise Sampling and Its Applications

引用

Journal of computer science & Technology 2015年第3期30卷 439-452页

作者：严冬明郭建伟王斌张晓鹏 Peter Wonk Visual Computing Center King Abdullah University of Science and Technology Thuwal 23955-6900 Saudi Arabia National Laboratory of Pattern Recognition Institute of Automation Chinese Academy of Sciences Beijing 100190 China School of Software Tsinghua University Beijing 100084 China Department of Computer Science and Engineering Arizona State University Tempe AZ 85287 U.S.A.

In this paper, we survey recent approaches to blue-noise sampling and discuss their beneficial applications. We discuss the sampling algorithms that use points as sampling primitives and classify the sampling algorithms based on various aspects, e.g., the sampling domain and the type of algorithm. We demonstrate several well-known applications that can be improved by recent blue-noise sampling techniques, as well as some new applications such as dynamic sampling and blue-noise remeshing.

关键词： blue-noise sampling Poisson-disk sampling Lloyd relaxation rendering remeshing

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：