CASIA OpenIR
Deep Crisp Boundaries: From Boundaries to Higher-level Tasks
Wang, Yupei; Zhao, Xin; Li, Yin; Huang, Kaiqi
Source PublicationIEEE Transactions on Image Processing
2019
Volume28Issue:3Pages:1285-1298
Abstract

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.

KeywordBoundary Detection, Deep Learning
Indexed BySCI
Document Type期刊论文
Identifierhttp://ir.ia.ac.cn/handle/173211/23349
Collection中国科学院自动化研究所
Affiliation1.Institute of Automation, Chinese Academy of Sciences
2.University of Chinese Academy of Sciences
Recommended Citation
GB/T 7714
Wang, Yupei,Zhao, Xin,Li, Yin,et al. Deep Crisp Boundaries: From Boundaries to Higher-level Tasks[J]. IEEE Transactions on Image Processing,2019,28(3):1285-1298.
APA Wang, Yupei,Zhao, Xin,Li, Yin,&Huang, Kaiqi.(2019).Deep Crisp Boundaries: From Boundaries to Higher-level Tasks.IEEE Transactions on Image Processing,28(3),1285-1298.
MLA Wang, Yupei,et al."Deep Crisp Boundaries: From Boundaries to Higher-level Tasks".IEEE Transactions on Image Processing 28.3(2019):1285-1298.
Files in This Item: Download All
File Name/Size DocType Version Access License
08485388(1).pdf(4781KB)期刊论文作者接受稿开放获取CC BY-NC-SAView Download
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Wang, Yupei]'s Articles
[Zhao, Xin]'s Articles
[Li, Yin]'s Articles
Baidu academic
Similar articles in Baidu academic
[Wang, Yupei]'s Articles
[Zhao, Xin]'s Articles
[Li, Yin]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Wang, Yupei]'s Articles
[Zhao, Xin]'s Articles
[Li, Yin]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: 08485388(1).pdf
Format: Adobe PDF
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.