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. 智能感知与计算研究中心
Incremental Generative Occlusion Adversarial Suppression Network for Person ReID

文献类型：期刊论文

作者Zhao, Cairong4; Lv, Xinbi4; Dou, Shuguang4; Zhang, Shanshan2,3; Wu, Jun4; Wang, Liang1
刊名IEEE TRANSACTIONS ON IMAGE PROCESSING
出版日期2021
卷号30页码:4212-4224
ISSN号1057-7149
关键词Feature extraction Training Image reconstruction Body regions Training data Cameras Two dimensional displays Batch-based incremental occlusion occlusion suppression occluded person re-identification
DOI10.1109/TIP.2021.3070182
通讯作者Zhao, Cairong(zhaocairong@tongji.edu.cn)
英文摘要Person re-identification (re-id) suffers from the significant challenge of occlusion, where an image contains occlusions and less discriminative pedestrian information. However, certain work consistently attempts to design complex modules to capture implicit information (including human pose landmarks, mask maps, and spatial information). The network, consequently, focuses on discriminative features learning on human non-occluded body regions and realizes effective matching under spatial misalignment. Few studies have focused on data augmentation, given that existing single-based data augmentation methods bring limited performance improvement. To address the occlusion problem, we propose a novel Incremental Generative Occlusion Adversarial Suppression (IGOAS) network. It consists of 1) an incremental generative occlusion block, generating easy-to-hard occlusion data, that makes the network more robust to occlusion by gradually learning harder occlusion instead of hardest occlusion directly. And 2) a global-adversarial suppression (G&A) framework with a global branch and an adversarial suppression branch. The global branch extracts steady global features of the images. The adversarial suppression branch, embedded with two occlusion suppression module, minimizes the generated occlusion's response and strengthens attentive feature representation on human non-occluded body regions. Finally, we get a more discriminative pedestrian feature descriptor by concatenating two branches' features, which is robust to the occlusion problem. The experiments on the occluded dataset show the competitive performance of IGOAS. On Occluded-DukeMTMC, it achieves 60.1% Rank-1 accuracy and 49.4% mAP.
资助项目National Natural Science Foundation of China (NSFC)[62076184] ; National Natural Science Foundation of China (NSFC)[61673299] ; National Natural Science Foundation of China (NSFC)[61976160] ; National Natural Science Foundation of China (NSFC)[61573255] ; Key Laboratory of Advanced Theory and Application in Statistics and Data Science, East China Normal University, Ministry of Education ; Fundamental Research Funds for the Central Universities
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:000640713600004
资助机构National Natural Science Foundation of China (NSFC) ; Key Laboratory of Advanced Theory and Application in Statistics and Data Science, East China Normal University, Ministry of Education ; Fundamental Research Funds for the Central Universities
源URL[http://ir.ia.ac.cn/handle/173211/44485]  
专题自动化研究所_智能感知与计算研究中心
通讯作者Zhao, Cairong
作者单位1.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit NLPR, Beijing 100190, Peoples R China
2.Nanjing Univ Sci & Technol, Jiangsu Key Lab Image & Video Understanding Socia, Nanjing 210094, Peoples R China
3.Nanjing Univ Sci & Technol, Sch Comp Sci & Engn, Nanjing 210094, Peoples R China
4.Tongji Univ, Dept Comp Sci & Technol, Shanghai 201804, Peoples R China
推荐引用方式
GB/T 7714		 Zhao, Cairong,Lv, Xinbi,Dou, Shuguang,et al. Incremental Generative Occlusion Adversarial Suppression Network for Person ReID[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING,2021,30:4212-4224.
APA Zhao, Cairong,Lv, Xinbi,Dou, Shuguang,Zhang, Shanshan,Wu, Jun,&Wang, Liang.(2021).Incremental Generative Occlusion Adversarial Suppression Network for Person ReID.IEEE TRANSACTIONS ON IMAGE PROCESSING,30,4212-4224.
MLA Zhao, Cairong,et al."Incremental Generative Occlusion Adversarial Suppression Network for Person ReID".IEEE TRANSACTIONS ON IMAGE PROCESSING 30(2021):4212-4224.

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。