Learning pseudo labels for semi-and-weakly supervised semantic segmentation

doi:10.1016/j.patcog.2022.108925

	Learning pseudo labels for semi-and-weakly supervised semantic segmentation
	Wang, Yude 1,2; Zhang, Jie 1,2; Kan, Meina 1,2; Shan, Shiguang 1,2,3
	2022-12-01
发表期刊	PATTERN RECOGNITION
ISSN	0031-3203
卷号	132 页码:10
摘要	In this paper, we aim to tackle semi-and-weakly supervised semantic segmentation (SWSSS), where many image-level classification labels and a few pixel-level annotations are available. We believe the most crucial point for solving SWSSS is to produce high-quality pseudo labels, and our method deals with it from two perspectives. Firstly, we introduce a class-aware cross entropy (CCE) loss for network training. Compared to conventional cross entropy loss, CCE loss encourages the model to distinguish concurrent classes only and simplifies the learning target of pseudo label generation. Secondly, we propose a progressive cross training (PCT) method to build cross supervision between two networks with a dynamic evaluation mechanism, which progressively introduces high-quality predictions as additional supervision for network training. Our method significantly improves the quality of generated pseudo labels in the regime with extremely limited annotations. Extensive experiments demonstrate that our approach outperforms state-of-the-art methods significantly. The code is released for public access 1 .(c) 2022 Elsevier Ltd. All rights reserved.
关键词	Semi -supervised Weakly supervised Semi -and -weakly supervised Semantic segmentation Pseudo label Self-training
DOI	10.1016/j.patcog.2022.108925
收录类别	SCI
语种	英语
资助项目	National Key R&D Program of China[2017YFA070 080 0] ; National Natural Science Founda- tion of China[62176251] ; National Natural Science Founda- tion of China[61976219] ; Beijing Nova Program[Z19110 0 0 01119123]
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Artificial Intelligence ; Engineering, Electrical & Electronic
WOS记录号	WOS:000874844700007
出版者	ELSEVIER SCI LTD
引用统计	被引频次：9[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/19755
专题	中国科学院计算技术研究所期刊论文
通讯作者	Zhang, Jie
作者单位	1.Chinese Acad Sci, Key Lab Intelligent Informat Proc Chinese Acad Sci, Inst Comp Technol, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Beijing 100049, Peoples R China 3.Peng Cheng Natl Lab, Shenzhen 518055, Guangdong, Peoples R China
推荐引用方式 GB/T 7714	Wang, Yude,Zhang, Jie,Kan, Meina,et al. Learning pseudo labels for semi-and-weakly supervised semantic segmentation[J]. PATTERN RECOGNITION,2022,132:10.
APA	Wang, Yude,Zhang, Jie,Kan, Meina,&Shan, Shiguang.(2022).Learning pseudo labels for semi-and-weakly supervised semantic segmentation.PATTERN RECOGNITION,132,10.
MLA	Wang, Yude,et al."Learning pseudo labels for semi-and-weakly supervised semantic segmentation".PATTERN RECOGNITION 132(2022):10.