Camera-Invariant Meta-Learning Network for Single-Camera-Training Person Reidentification

doi:10.1109/JIOT.2025.3550976

	Camera-Invariant Meta-Learning Network for Single-Camera-Training Person Reidentification
	Pei, Jiangbo 1,2; Jiang, Zhuqing 1,2; Men, Aidong 1,2; Wang, Haiying 1,2; Luo, Haiyong 3; Wen, Shiping 4
	2025-06-15
发表期刊	IEEE INTERNET OF THINGS JOURNAL
ISSN	2327-4662
卷号	12 期号:12 页码:22381-22392
摘要	Single-camera-training person reidentification (SCT re-ID) aims to train a reidentification (re-ID) model using single-camera-training (SCT) datasets where each person appears in only one camera. The main challenge of SCT re-ID is to learn camera-invariant feature representations without cross-camera same-person (CCSP) data as supervision. Previous methods address it by assuming that the most similar person should be found in another camera. However, this assumption is not guaranteed to be correct. In this article, we propose a novel solution: the camera-invariant meta-learning network (CIMN) for SCT re-ID. CIMN operates under the premise that camera-invariant feature representations should remain robust despite changes in camera settings. To achieve this, we partition the training data into a meta-train set and a meta-test set based on camera IDs. We then conduct a cross-camera simulation (CCS) using a meta-learning strategy, aiming to enforce the feature representations learned from the meta-train set to be robust when applied to the meta-test set. We further introduce three specific loss functions to leverage potential identity relations between the meta-train set and the meta-test set. Through the CCS and the introduced loss functions, CIMN can extract feature representations that are both camera-invariant and identity-discriminative even in the absence of CCSP data. Our experimental results demonstrate that CIMN can extract feature representations that are both camera-invariant and identity-discriminative, even in the absence of CCSP data. our method achieves comparable performance with and without the use of CCSP data, and outperforms state-of-the-art methods on three SCT re-ID benchmarks.
关键词	Cameras Training Metalearning Annotations Feature extraction Data models Identification of persons Costs Data mining Measurement Camera-invariant features meta-learning person reidentification (re-ID) single-camera-training (SCT)
DOI	10.1109/JIOT.2025.3550976
收录类别	SCI
语种	英语
资助项目	National Key Research and Development Program[2018YFB0505200] ; National Natural Science Funding[62002026]
WOS研究方向	Computer Science ; Engineering ; Telecommunications
WOS类目	Computer Science, Information Systems ; Engineering, Electrical & Electronic ; Telecommunications
WOS记录号	WOS:001506725100001
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/42294
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Jiang, Zhuqing
作者单位	1.Beijing Univ Posts & Telecommun, Sch Artificial Intelligence, Beijing 100876, Peoples R China 2.Beijing Univ Posts & Telecommun, Beijing Key Lab Network Syst & Network Culture, Beijing 100876, Peoples R China 3.Chinese Acad Sci, Inst Comp Technol, Beijing 100190, Peoples R China 4.Univ Technol Sydney, Australian AI Inst, Fac Engn & Informat Technol, Ultimo, NSW 2007, Australia
推荐引用方式 GB/T 7714	Pei, Jiangbo,Jiang, Zhuqing,Men, Aidong,et al. Camera-Invariant Meta-Learning Network for Single-Camera-Training Person Reidentification[J]. IEEE INTERNET OF THINGS JOURNAL,2025,12(12):22381-22392.
APA	Pei, Jiangbo,Jiang, Zhuqing,Men, Aidong,Wang, Haiying,Luo, Haiyong,&Wen, Shiping.(2025).Camera-Invariant Meta-Learning Network for Single-Camera-Training Person Reidentification.IEEE INTERNET OF THINGS JOURNAL,12(12),22381-22392.
MLA	Pei, Jiangbo,et al."Camera-Invariant Meta-Learning Network for Single-Camera-Training Person Reidentification".IEEE INTERNET OF THINGS JOURNAL 12.12(2025):22381-22392.