Learning Domain Invariant Representations for Generalizable Person Re-Identification

doi:10.1109/TIP.2022.3229621

CASIA OpenIR > 智能感知与计算研究中心

	Learning Domain Invariant Representations for Generalizable Person Re-Identification
	Zhang, Yi-Fan1,2 ; Zhang, Zhang1,2 ; Li, Da 1,2; Jia, Zhen1,2 ; Wang, Liang3,4,5 ; Tan, Tieniu3,4,5
发表期刊	IEEE TRANSACTIONS ON IMAGE PROCESSING
ISSN	1057-7149
	2023
卷号	32 页码:509-523
通讯作者	Zhang, Zhang(zzhang@nlpr.ia.ac.cn)
摘要	Generalizable person Re-Identification (ReID) aims to learn ready-to-use cross-domain representations for direct cross-data evaluation, which has attracted growing attention in the recent computer vision (CV) community. In this work, we construct a structural causal model (SCM) among identity labels, identity-specific factors (clothing/shoes color etc.), and domain-specific factors (background, viewpoints etc.). According to the causal analysis, we propose a novel Domain Invariant Representation Learning for generalizable person Re-Identification (DIR-ReID) framework. Specifically, we propose to disentangle the identity-specific and domain-specific factors into two independent feature spaces, based on which an effective backdoor adjustment approximate implementation is proposed for serving as a causal intervention towards the SCM. Extensive experiments have been conducted, showing that DIR-ReID outperforms state-of-the-art (SOTA) methods on large-scale domain generalization (DG) ReID benchmarks.
关键词	Generalizable person re-Identification disentanglement backdoor adjustment
DOI	10.1109/TIP.2022.3229621
关键词[WOS]	NETWORK
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China[62106260] ; National Natural Science Foundation of China[62236010] ; National Natural Science Foundation of China[61721004] ; National Natural Science Foundation of China[U1803261] ; China Postdoctoral Science Foundation[2020M680751]
项目资助者	National Natural Science Foundation of China ; China Postdoctoral Science Foundation
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Artificial Intelligence ; Engineering, Electrical & Electronic
WOS记录号	WOS:000906943400004
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计	被引频次：7[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/51348
专题	智能感知与计算研究中心多模态人工智能系统全国重点实验室
通讯作者	Zhang, Zhang
作者单位	1.Chinese Acad Sci CASIA, Inst Automat, Ctr Res Intelligent Percept & Comp CRIPAC, Natl Lab Pattern Recognit NLPR, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci UCAS, Sch Artificial Intelligence, Beijing 100049, Peoples R China 3.Chinese Acad Sci CASIA, Inst Automat, Ctr Res Intelligent Percept & Comp, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China 4.Univ Chinese Acad Sci, Beijing 100044, Peoples R China 5.Univ Chinese Acad Sci UCAS, Sch Artificial Intelligence, Beijing 100190, Peoples R China
第一作者单位	模式识别国家重点实验室
通讯作者单位	模式识别国家重点实验室
推荐引用方式 GB/T 7714	Zhang, Yi-Fan,Zhang, Zhang,Li, Da,et al. Learning Domain Invariant Representations for Generalizable Person Re-Identification[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING,2023,32:509-523.
APA	Zhang, Yi-Fan,Zhang, Zhang,Li, Da,Jia, Zhen,Wang, Liang,&Tan, Tieniu.(2023).Learning Domain Invariant Representations for Generalizable Person Re-Identification.IEEE TRANSACTIONS ON IMAGE PROCESSING,32,509-523.
MLA	Zhang, Yi-Fan,et al."Learning Domain Invariant Representations for Generalizable Person Re-Identification".IEEE TRANSACTIONS ON IMAGE PROCESSING 32(2023):509-523.