An end-to-end exemplar association for unsupervised person Re-identification

doi:10.1016/j.neunet.2020.05.015

CORC > 自动化研究所 > 中国科学院自动化研究所 > 模式识别国家重点实验室 > 图像与视频分析团队

	An end-to-end exemplar association for unsupervised person Re-identification
	Wu, Jinlin 1,4,5; Yang, Yang 1,4,5; Lei, Zhen 1,4,5; Wang, Jinqiao 1,4,5; Li, Stan Z.3; Tiwari, Prayag 6; Pandey, Hari Mohan 2
刊名	NEURAL NETWORKS
	2020-09-01
卷号	129 页码:43-54
关键词	End-to-end exemplar-based training Exemplar association Dynamic selection threshold
ISSN号	0893-6080
DOI	10.1016/j.neunet.2020.05.015
通讯作者	Lei, Zhen(zlei@nlpr.ia.ac.cn)
英文摘要	Tracklet association methods learn the cross camera retrieval ability though associating underlying cross camera positive samples, which have proven to be successful in unsupervised person re-identification task. However, most of them use poor-efficiency association strategies which costs long training hours but gains the low performance. To solve this, we propose an effective end-to-end exemplar associations (EEA) framework in this work. EEA mainly adapts three strategies to improve efficiency: (1) end-to-end exemplar-based training, (2) exemplar association and (3) dynamic selection threshold. The first one is to accelerate the training process, while the others aim to improve the tracklet association precision. Compared with existing tracklet associating methods, EEA obviously reduces the training cost and achieves the higher performance. Extensive experiments and ablation studies on seven RE-ID datasets demonstrate the superiority of the proposed EEA over most state-of-the-art unsupervised and domain adaptation RE-ID methods. (C) 2020 Elsevier Ltd. All rights reserved.
资助项目	National Key Research and Development Plan[2019YFC2003901] ; Chinese National Natural Science Foundation[61806203] ; Chinese National Natural Science Foundation[61876178] ; Chinese National Natural Science Foundation[61872367] ; Chinese National Natural Science Foundation[61976229]
WOS研究方向	Computer Science ; Neurosciences & Neurology
语种	英语
出版者	PERGAMON-ELSEVIER SCIENCE LTD
WOS记录号	WOS:000555927200005
资助机构	National Key Research and Development Plan ; Chinese National Natural Science Foundation
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/40344]
专题	自动化研究所_模式识别国家重点实验室_图像与视频分析团队
通讯作者	Lei, Zhen
作者单位	1.Chinese Acad Sci, Inst Automat, CBSR, Beijing, Peoples R China 2.Edge Hill Univ, Dept Comp Sci, Ormskirk, England 3.Westlake Univ, Sch Engn, Hangzhou, Peoples R China 4.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing, Peoples R China 5.Chinese Acad Sci, Inst Automat, NLPR, Beijing, Peoples R China 6.Univ Padua, Dept Informat Engn, Padua, Italy
推荐引用方式 GB/T 7714	Wu, Jinlin,Yang, Yang,Lei, Zhen,et al. An end-to-end exemplar association for unsupervised person Re-identification[J]. NEURAL NETWORKS,2020,129:43-54.
APA	Wu, Jinlin.,Yang, Yang.,Lei, Zhen.,Wang, Jinqiao.,Li, Stan Z..,...&Pandey, Hari Mohan.(2020).An end-to-end exemplar association for unsupervised person Re-identification.NEURAL NETWORKS,129,43-54.
MLA	Wu, Jinlin,et al."An end-to-end exemplar association for unsupervised person Re-identification".NEURAL NETWORKS 129(2020):43-54.