Siamese Regression Tracking With Reinforced Template Updating

doi:10.1109/TIP.2020.3036723

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

	Siamese Regression Tracking With Reinforced Template Updating
	Zhao, Fei 2,7; Zhang, Ting 3; Song, Yibing 1; Tang, Ming 2,4; Wang, Xiaobo 7; Wang, Jinqiao 2,5,6
刊名	IEEE TRANSACTIONS ON IMAGE PROCESSING
	2021
卷号	30 页码:628-640
关键词	Target tracking Training Reinforcement learning Visualization Task analysis Benchmark testing Head Siamese regression tracking actor-critic network reinforcement learning
ISSN号	1057-7149
DOI	10.1109/TIP.2020.3036723
通讯作者	Wang, Jinqiao(jqwang@nlpr.ia.ac.cn)
英文摘要	Siamese networks are prevalent in visual tracking because of the efficient localization. The networks take both a search patch and a target template as inputs where the target template is usually from the initial frame. Meanwhile, Siamese trackers do not update network parameters online for real-time efficiency. The fixed target template and CNN parameters make Siamese trackers not effective to capture target appearance variations. In this paper, we propose a template updating method via reinforcement learning for Siamese regression trackers. We collect a series of templates and learn to maintain them based on an actor-critic framework. Among this framework, the actor network that is trained by deep reinforcement learning effectively updates the templates based on the tracking result on each frame. Besides the target template, we update the Siamese regression tracker online to adapt to target appearance variations. The experimental results on the standard benchmarks show the effectiveness of both template and network updating. The proposed tracker SiamRTU performs favorably against state-of-the-art approaches.
资助项目	Research and Development Projects in the Key Areas of Guangdong Province[2020B010165001] ; National Natural Science Foundation of China[61772527] ; National Natural Science Foundation of China[61976210] ; National Natural Science Foundation of China[61806200] ; National Natural Science Foundation of China[61702510] ; National Natural Science Foundation of China[61876086] ; National Natural Science Foundation of China[62076235]
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:000597161500002
资助机构	Research and Development Projects in the Key Areas of Guangdong Province ; National Natural Science Foundation of China
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/42664]
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	Wang, Jinqiao
作者单位	1.Tencent AI Lab, Shenzhen 518172, Peoples R China 2.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China 3.CEIEC, Res & Dev Ctr, Beijing 100036, Peoples R China 4.Shenzhen Infinova Ltd, Shenzhen 518110, Peoples R China 5.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing 100049, Peoples R China 6.ObjectEye Inc, Beijing 100080, Peoples R China 7.Alibaba Grp, Hangzhou 311100, Peoples R China
推荐引用方式 GB/T 7714	Zhao, Fei,Zhang, Ting,Song, Yibing,et al. Siamese Regression Tracking With Reinforced Template Updating[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING,2021,30:628-640.
APA	Zhao, Fei,Zhang, Ting,Song, Yibing,Tang, Ming,Wang, Xiaobo,&Wang, Jinqiao.(2021).Siamese Regression Tracking With Reinforced Template Updating.IEEE TRANSACTIONS ON IMAGE PROCESSING,30,628-640.
MLA	Zhao, Fei,et al."Siamese Regression Tracking With Reinforced Template Updating".IEEE TRANSACTIONS ON IMAGE PROCESSING 30(2021):628-640.