JWSAA: Joint weak saliency and attention aware for person re-identification

被引：125
作者：
Ning, Xin [1 ,2 ,3 ,5 ,6 ]
Gong, Ke [5 ,6 ]
Li, Weijun [1 ,2 ,3 ,4 ,5 ]
Zhang, Liping [1 ,5 ,6 ]
机构：
[1] Chinese Acad Sci, Inst Semicond, Beijing 100083, Peoples R China
[2] Univ Chinese Acad Sci, Ctr Mat Sci & Optoelect Engn, Beijing 100049, Peoples R China
[3] Univ Chinese Acad Sci, Sch Microelect, Beijing 100049, Peoples R China
[4] Beijing Key Lab Semicond Neural Network Intellige, Beijing 100083, Peoples R China
[5] Beijing Wave Secur Technol Co Ltd, Beijing 102208, Peoples R China
[6] Wave Grp, Cognit Comp Technol Joint Lab, Beijing 100083, Peoples R China
来源：
NEUROCOMPUTING | 2021年 / 453卷
基金：
中国国家自然科学基金;
关键词：
Deep neural networks; Attention mechanism; Saliency features; Person re-identification;
D O I：
10.1016/j.neucom.2020.05.106
中图分类号：
TP18 [人工智能理论];
学科分类号：
081104 ; 0812 ; 0835 ; 1405 ;
摘要：
Attention mechanisms can extract salient features in images, which has been proven to be effective for person re-identification. However, focusing on the saliency of an image is not enough. On the one hand, the salient features extracted from the model are not necessarily the features needed, e.g., a similar background may also be mistaken as salient features; on the other hand, various salient features are often more conducive to improving the performance of the model. Based on this, in this paper, a model that has joint weak saliency and attention aware is proposed, which can obtain more complete global features by weakening saliency features. The model then obtains diversified saliency features via attention diversity to improve the performance of the model. Experiments on commonly used datasets prove the effectiveness of the proposed method. (c) 2020 Elsevier B.V. All rights reserved.
引用
收藏
页码：801 / 811
页数：11
相关论文
共 60 条
[1]  
Bolle R.M., 2005, Fourth IEEEWorkshop on Automatic Identification Advanced Technologies, DOI [DOI 10.1109/AUTOID.2005.48, 10.1109/AUTOID.2005.48]
[2]  
Chen HR, 2018, 2018 IEEE FOURTH INTERNATIONAL CONFERENCE ON MULTIMEDIA BIG DATA (BIGMM)
[3]   ABD-Net: Attentive but Diverse Person Re-Identification [J].
Chen, Tianlong ;
Ding, Shaojin ;
Xie, Jingyi ;
Yuan, Ye ;
Chen, Wuyang ;
Yang, Yang ;
Ren, Zhou ;
Wang, Zhangyang .
2019 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2019), 2019, :8350-8360
[4]   Person Re-Identification by Multi-Channel Parts-Based CNN with Improved Triplet Loss Function [J].
Cheng, De ;
Gong, Yihong ;
Zhou, Sanping ;
Wang, Jinjun ;
Zheng, Nanning .
2016 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2016, :1335-1344
[5]  
Deng J, 2009, PROC CVPR IEEE, P248, DOI 10.1109/CVPRW.2009.5206848
[6]  
Felzenszwalb P, 2008, PROC CVPR IEEE, P1984
[7]   Attention Branch Network: Learning of Attention Mechanism for Visual Explanation [J].
Fukui, Hiroshi ;
Hirakawa, Tsubasa ;
Yamashita, Takayoshi ;
Fujiyoshi, Hironobu .
2019 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2019), 2019, :10697-10706
[8]  
Guo H., 2019, 100000 KEY LAB 2019
[9]   Deep Residual Learning for Image Recognition [J].
He, Kaiming ;
Zhang, Xiangyu ;
Ren, Shaoqing ;
Sun, Jian .
2016 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2016, :770-778
[10]  
Hermans Alexander, 2017, Defense of the Triplet loss
123456