中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
首页 机构 成果 学者
  
  1. CAS IR Grid
  2. 自动化研究所
  3. 中国科学院自动化研究所
  4. 多模态人工智能系统全国重点实验室
Jointing Recurrent Across-Channel and Spatial Attention for Multi-Object Tracking With Block-Erasing Data Augmentation

文献类型：期刊论文

作者Deng, Keyu1; Zhang, Congxuan2,4; Chen, Zhen2,4; Hu, Weiming3; Li, Bing3; Lu, Feng2,4
刊名IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
出版日期2023-08-01
卷号33期号:8页码:4054-4069
ISSN号1051-8215
关键词Multi-object tracking one shot multiattention feature learning block erasing strategy object occlusions
DOI10.1109/TCSVT.2023.3238716
通讯作者Zhang, Congxuan(zcxdsg@163.com) ; Chen, Zhen(dr_chenzhen@163.com)
英文摘要Although deep-learning-based multi-object tracking (MOT) approaches have achieved remarkable performances in terms of accuracy and efficiency, the issue of object occlusions remains an open challenge for most one-shot MOT methods. To address the problem of object occlusions, in this paper we present a recurrent across-channel and spatial attention-based one-shot multi-object tracking method with block-erasing data augmentation. First, we construct a multiattention feature learning module, named RASFL, that combines recurrent across -channel attention with spatial attention. The RASFL extracts both the correlations of the feature channels and the differences of the spatial locations to improve the accuracy of the re-identification (Re-ID) task. Second, we adopt a block-erasing data augmentation strategy to handle object occlusions by using random pixel blocks to simulate occlusion cases during the network training process. This block-erasing data augmentation assists the network to be more robust under object occlusions. By integrating the proposed RASFL module and the block-erasing data augmentation strategy into a one-shot online MOT system, we build an accurate and robust MOT model called DcMOT. Finally, we run our method on the MOT16, MOT17 and MOT20 datasets to conduct a comprehensive comparison with some of the state-of-the-art MOT methods. The experimental results demonstrate that the proposed DcMOT model achieves a competitive performance in terms of both accuracy and efficiency; with especially good performances in the occlusion cases.
WOS关键词OBJECT TRACKING ; REIDENTIFICATION ; MULTITARGET ; SYSTEM
资助项目National Key Research and Development Program of China[2020YFC2003800] ; National Natural Science Foundation of China[62222206] ; National Natural Science Foundation of China[62272209] ; National Natural Science Foundation of China[61866026] ; National Natural Science Foundation of China[61866025] ; National Natural Science Foundation of Jiangxi Province[20202ACB214007] ; Technology Innovation Guidance Program of Jiangxi Province[20212AEI91005]
WOS研究方向Engineering
语种英语
出版者IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
WOS记录号WOS:001045167400040
资助机构National Key Research and Development Program of China ; National Natural Science Foundation of China ; National Natural Science Foundation of Jiangxi Province ; Technology Innovation Guidance Program of Jiangxi Province
源URL[http://ir.ia.ac.cn/handle/173211/54032]  
专题多模态人工智能系统全国重点实验室
通讯作者Zhang, Congxuan; Chen, Zhen
作者单位1.Nanchang Hangkong Univ, Sch Measuring & Opt Engn, Nanchang 330063, Peoples R China
2.Nanchang Hangkong Univ, Key Lab Nondestruct Testing, Minist Educ, Nanchang 330063, Peoples R China
3.Chinese Acad Sci, Inst Automat, Beijing 100190, Peoples R China
4.Nanchang Hangkong Univ, Sch Measuring & Opt Engn, Nanchang 330063, Peoples R China
推荐引用方式
GB/T 7714		 Deng, Keyu,Zhang, Congxuan,Chen, Zhen,et al. Jointing Recurrent Across-Channel and Spatial Attention for Multi-Object Tracking With Block-Erasing Data Augmentation[J]. IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY,2023,33(8):4054-4069.
APA Deng, Keyu,Zhang, Congxuan,Chen, Zhen,Hu, Weiming,Li, Bing,&Lu, Feng.(2023).Jointing Recurrent Across-Channel and Spatial Attention for Multi-Object Tracking With Block-Erasing Data Augmentation.IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY,33(8),4054-4069.
MLA Deng, Keyu,et al."Jointing Recurrent Across-Channel and Spatial Attention for Multi-Object Tracking With Block-Erasing Data Augmentation".IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY 33.8(2023):4054-4069.

入库方式： OAI收割

来源：自动化研究所

浏览0
下载0
收藏0
其他版本

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。