Contour-Aware Equipotential Learning for Semantic Segmentation
- 주제(키워드) Category-level contour learning , semantic boundary refinement , supervised semantic segmentation
- 주제(기타) Computer Science, Information Systems; Computer Science, Software Engineering; Telecommunications
- 설명문(일반) [Yin, Xu] Korea Adv Inst Sci & Technol, Sch Comp, Daejeon 34141, South Korea; [Min, Dongbo] Ewha Womans Univ, Dept Comp Sci & Engn, Seoul 03760, South Korea; [Huo, Yuchi] Zhejiang Lab, Hangzhou 310058, Peoples R China; [Huo, Yuchi] Zhejiang Univ, State Key Lab CAD & CG, Hangzhou 310058, Peoples R China; [Yoon, Sung-Eui] Korea Adv Inst Sci & Technol, Fac Sch Comp, Daejeon 34141, South Korea
- 관리정보기술 faculty
- 등재 SCIE, SCOPUS
- OA유형 Green Submitted
- 발행기관 IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
- 발행년도 2023
- 총서유형 Journal
- URI http://www.dcollection.net/handler/ewha/000000213722
- 본문언어 영어
- Published As https://doi.org/10.1109/TMM.2022.3205441
초록/요약
With increasing demands for high-quality semantic segmentation in the industry, hard-distinguishing semantic boundaries have posed a significant threat to existing solutions. Inspired by real-life experience, i.e., combining varied observations contributes to higher visual recognition confidence, we present the equipotential learning (EPL) method. This novel module transfers the predicted/ground-truth semantic labels to a self-defined potential domain to learn and infer decision boundaries along customized directions. The conversion to the potential domain is implemented via a lightweight differentiable anisotropic convolution without incurring any parameter overhead. Besides, the designed two loss functions, the point loss and the equipotential line loss implement anisotropic field regression and category-level contour learning, respectively, enhancing prediction consistencies in the inter/intra-class boundary areas. More importantly, EPL is agnostic to network architectures, and thus it can be plugged into most existing segmentation models. This paper is the first attempt to address the boundary segmentation problem with field regression and contour learning. Meaningful performance improvements on Pascal Voc 2012 and Cityscapes demonstrate that the proposed EPL module can benefit the off-the-shelf fully convolutional network models when recognizing semantic boundary areas. Besides, intensive comparisons and analysis show the favorable merits of EPL for distinguishing semantically-similar and irregular-shaped categories.
more