Feature Aggregation With Reinforcement Learning for Video-Based Person Re-Identification

doi:10.1109/TNNLS.2019.2899588

CORC > 自动化研究所 > 中国科学院自动化研究所 > 复杂系统管理与控制国家重点实验室 > 机器人应用与理论组

	Feature Aggregation With Reinforcement Learning for Video-Based Person Re-Identification
	Zhang, Wei 2; He, Xuanyu 2; Lu, Weizhi 2; Qiao, Hong 1; Li, Yibin 2
刊名	IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
	2019-12-01
卷号	30 期号:12 页码:3847-3852
关键词	Feature extraction Task analysis Cameras Noise measurement Learning systems Reinforcement learning Feature aggregation reinforcement learning (RL) sequential decision making video-based person re-identification (re-id)
ISSN号	2162-237X
DOI	10.1109/TNNLS.2019.2899588
通讯作者	He, Xuanyu(hexiffer@outlook.com)
英文摘要	Video-based person re-identification (re-id) matches two tracks of persons from different cameras. Features are extracted from the images of a sequence and then aggregated as a track feature. Compared to existing works that aggregate frame features by simply averaging them or using temporal models such as recurrent neural networks, we propose an intelligent feature aggregate method based on reinforcement learning. Specifically, we train an agent to determine which frames in the sequence should be abandoned in the aggregation, which can be treated as a decision making process. By this way, the proposed method avoids introducing noisy information of the sequence and retains these valuable frames when generating a track feature. On benchmark data sets, experimental results show that our method can boost the re-id accuracy obviously based on the state-of-the-art models.
资助项目	National Key Research and Development Plan of China[2017YFB1300205] ; NSFC[61573222] ; NSFC[61801264] ; Major Research Program of Shandong Province[2018CXGC1503] ; Fundamental Research Funds of Shandong University[2016JC014] ; Basic Research Program of Shenzhen[JCYJ20170307153635551]
WOS研究方向	Computer Science ; Engineering
语种	英语
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
WOS记录号	WOS:000502762600028
资助机构	National Key Research and Development Plan of China ; NSFC ; Major Research Program of Shandong Province ; Fundamental Research Funds of Shandong University ; Basic Research Program of Shenzhen
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/29439]
专题	自动化研究所_复杂系统管理与控制国家重点实验室_机器人应用与理论组
通讯作者	He, Xuanyu
作者单位	1.Chinese Acad Sci, Inst Automat, Beijing 100190, Peoples R China 2.Shandong Univ, Sch Control Sci & Engn, Jinan 250100, Shandong, Peoples R China
推荐引用方式 GB/T 7714	Zhang, Wei,He, Xuanyu,Lu, Weizhi,et al. Feature Aggregation With Reinforcement Learning for Video-Based Person Re-Identification[J]. IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS,2019,30(12):3847-3852.
APA	Zhang, Wei,He, Xuanyu,Lu, Weizhi,Qiao, Hong,&Li, Yibin.(2019).Feature Aggregation With Reinforcement Learning for Video-Based Person Re-Identification.IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS,30(12),3847-3852.
MLA	Zhang, Wei,et al."Feature Aggregation With Reinforcement Learning for Video-Based Person Re-Identification".IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 30.12(2019):3847-3852.