Pseudo Segmentation for Semantic Information-Aware Stereo Matching

doi:10.1109/LSP.2022.3158586

	Pseudo Segmentation for Semantic Information-Aware Stereo Matching
	Hua, Shengyou 1,2; Sun, Zhiyong 1; Song, Bo 1; Liang, Pengpeng 3; Cheng, Erkang 1
刊名	IEEE SIGNAL PROCESSING LETTERS
	2022
卷号	29
关键词	Costs Semantics Feature extraction Image segmentation Task analysis Three-dimensional displays Correlation Deep learning stereo matching pseudo segmenta- tion
ISSN号	1070-9908
DOI	10.1109/LSP.2022.3158586
通讯作者	Song, Bo(songbo@iim.ac.cn) ; Liang, Pengpeng(liangpcs@gmail.com)
英文摘要	Stereo matching plays an important role in computer vision and robotics. Though substantial progress has been made on deep learning-based algorithms, the inherent semantic information within the ground truth of the training data for stereo matching has not been well explored. In this letter, we propose to use a pseudo segmentation sub-network to extract additional semantic information. More specifically, we divide the disparity label into groups and let each group correspond to a class for pseudo segmentation. To assist stereo matching with the semantic information obtained from pseudo segmentation, we inject the feature maps at the end of the pseudo segmentation sub-network into the cost volume that is used to infer the pixel-level disparity. To validate the effectiveness of the proposed approach, we select PSMNet (Chang and Chen, 2018)and GwcNet (Guo et al., 2019) as baselines and enhance them with the pseudo segmentation sub-network. Comprehensive experiments are carried out on the Scene Flow, KITTI 2015, and KITTI 2012 datasets, and the results show that our proposed method can improve the performance notably.
资助项目	National Natural Science Foundation of China[61806181] ; National Natural Science Foundation of China[61973294] ; KRDP of Anhui Province[201904a05020086] ; CAS[GJTD-2018-15] ; University Synergy Innovation Program of Anhui Province[GXXT-2021-030]
WOS研究方向	Engineering
语种	英语
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
WOS记录号	WOS:000776043500009
资助机构	National Natural Science Foundation of China ; KRDP of Anhui Province ; CAS ; University Synergy Innovation Program of Anhui Province
内容类型	期刊论文
源URL	[http://ir.hfcas.ac.cn:8080/handle/334002/128207]
专题	中国科学院合肥物质科学研究院
通讯作者	Song, Bo; Liang, Pengpeng
作者单位	1.Chinese Acad Sci, HFIPS, Inst Intelligent Machines, Hefei 230031, Peoples R China 2.Univ Sci & Technol China, Hefei 230026, Peoples R China 3.Zhengzhou Univ, Sch Comp & Artificial Intelligence, Zhengzhou 450001, Peoples R China
推荐引用方式 GB/T 7714	Hua, Shengyou,Sun, Zhiyong,Song, Bo,et al. Pseudo Segmentation for Semantic Information-Aware Stereo Matching[J]. IEEE SIGNAL PROCESSING LETTERS,2022,29.
APA	Hua, Shengyou,Sun, Zhiyong,Song, Bo,Liang, Pengpeng,&Cheng, Erkang.(2022).Pseudo Segmentation for Semantic Information-Aware Stereo Matching.IEEE SIGNAL PROCESSING LETTERS,29.
MLA	Hua, Shengyou,et al."Pseudo Segmentation for Semantic Information-Aware Stereo Matching".IEEE SIGNAL PROCESSING LETTERS 29(2022).