Robust Visual Tracking via Hierarchical Particle Filter and Ensemble Deep Features

doi:10.1109/TCSVT.2018.2889457

	Robust Visual Tracking via Hierarchical Particle Filter and Ensemble Deep Features
	Li, Shengjie 1,2; Zhao, Shuai 1,2; Cheng, Bo 1,2; Zhao, Erhu 3; Chen, Junliang 1
	2020
发表期刊	IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
ISSN	1051-8215
卷号	30 期号:1 页码:179-191
摘要	Particle filter algorithms are a very important branch for visual object tracking in the past decades, showing strong robustness to challenging scenarios with partial occlusion and large-scale variations. However, since a large number of particles need to be extracted for the accurate target state estimation, their tracking efficiency typically suffers especially when meeting deep convolutional features, which have been developed for handling significant variations of the target appearance in the visual tracking community. In this paper, we propose to elegantly exploit deep convolutional features with few particles in a novel hierarchical particle filter, which formulates correlation filters as observation models and breaks the standard particle filter framework down into two constituent particle layers, namely, particle translation layer and particle scale layer. The particle translation layer focuses on the object location with the deep convolutional features capturing semantics but failing to precisely estimate the object scale, while the particle scale layer pays attention to large-scale variations with the lightweight hand-crafted features handling spatial details of the object size. Moreover, an efficient ensemble method is proposed to help explore deeper convolutional features with more semantics in the particle translation layer. Extensive experiments on four challenging tracking datasets, including OTB-2013, OTB-2015, VOT2014, and VOT2015 demonstrate that the proposed method performs favorably against a number of state-of-the-art trackers.
关键词	Object tracking correlation filter hierarchical particle filter convolutional neural networks
DOI	10.1109/TCSVT.2018.2889457
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China[61501048] ; Beijing Natural Science Foundation[4182042] ; Fundamental Research Funds for the Central Universities[2017RC12]
WOS研究方向	Engineering
WOS类目	Engineering, Electrical & Electronic
WOS记录号	WOS:000521641800015
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计	被引频次：16[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/14032
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Zhao, Shuai
作者单位	1.Beijing Univ Posts & Telecommun, State Key Lab Networking & Switching Technol, Beijing 100876, Peoples R China 2.Beijing Adv Innovat Ctr Future Internet Technol, Beijing 100124, Peoples R China 3.Chinese Acad Sci, Inst Comp Technol, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Li, Shengjie,Zhao, Shuai,Cheng, Bo,et al. Robust Visual Tracking via Hierarchical Particle Filter and Ensemble Deep Features[J]. IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY,2020,30(1):179-191.
APA	Li, Shengjie,Zhao, Shuai,Cheng, Bo,Zhao, Erhu,&Chen, Junliang.(2020).Robust Visual Tracking via Hierarchical Particle Filter and Ensemble Deep Features.IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY,30(1),179-191.
MLA	Li, Shengjie,et al."Robust Visual Tracking via Hierarchical Particle Filter and Ensemble Deep Features".IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY 30.1(2020):179-191.