Deep Crisp Boundaries: From Boundaries to Higher-Level Tasks

被引:59
作者
Wang, Yupei [1 ,2 ]
Zhao, Xin [1 ,2 ]
Li, Yin [3 ,4 ]
Huang, Kaiqi [5 ,6 ]
机构
[1] Chinese Acad Sci, Ctr Res Intelligent Syst & Engn, Inst Automat, Beijing 100190, Peoples R China
[2] Univ Chinese Aca Sci, Beijing 100049, Peoples R China
[3] Univ Wisconsin, Dept Biostat & Med Informat, Madison, WI 53706 USA
[4] Univ Wisconsin, Dept Comp Sci, Madison, WI 53706 USA
[5] Chinese Acad Sci, Natl Lab Pattern Recognit, Ctr Res Intelligent Syst & Engn, Inst Automat, Beijing, Peoples R China
[6] CAS Ctr Excellence Brain Sci & Intelligence Techn, Shanghai 200031, Peoples R China
基金
中国国家自然科学基金;
关键词
Boundary detection; deep learning;
D O I
10.1109/TIP.2018.2874279
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
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 the 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.
引用
收藏
页码:1285 / 1298
页数:14
相关论文
共 62 条
[1]  
[Anonymous], 2016, PROC CVPR IEEE, DOI DOI 10.1109/CVPR.2016.348
[2]  
[Anonymous], 2016, PROC CVPR IEEE, DOI DOI 10.1109/CVPR.2016.492
[3]   Multiscale Combinatorial Grouping [J].
Arbelaez, Pablo ;
Pont-Tuset, Jordi ;
Barron, Jonathan T. ;
Marques, Ferran ;
Malik, Jitendra .
2014 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2014, :328-335
[4]   Contour Detection and Hierarchical Image Segmentation [J].
Arbelaez, Pablo ;
Maire, Michael ;
Fowlkes, Charless ;
Malik, Jitendra .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2011, 33 (05) :898-916
[5]  
Bansal A., 2017, Pixelnet: Representation of the pixels, by the pixels, and for the pixels
[6]   Semantic Segmentation with Boundary Neural Fields [J].
Bertasius, Gedas ;
Shi, Jianbo ;
Torresani, Lorenzo .
2016 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2016, :3602-3610
[7]   High-for-Low and Low-for-High: Efficient Boundary Detection from Deep Object Features and its Applications to High-Level Vision [J].
Bertasius, Gedas ;
Shi, Jianbo ;
Torresani, Lorenzo .
2015 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2015, :504-512
[8]  
Bertasius G, 2015, PROC CVPR IEEE, P4380, DOI 10.1109/CVPR.2015.7299067
[9]   A Naturalistic Open Source Movie for Optical Flow Evaluation [J].
Butler, Daniel J. ;
Wulff, Jonas ;
Stanley, Garrett B. ;
Black, Michael J. .
COMPUTER VISION - ECCV 2012, PT VI, 2012, 7577 :611-625