Camera-Driven Representation Learning for Unsupervised Domain Adaptive Person Re-identification

被引:7
|
作者
Lee, Geon [1 ]
Lee, Sanghoon [1 ]
Kim, Dohyung [1 ]
Shin, Younghoon [2 ]
Yoon, Yongsang [2 ]
Ham, Bumsub [1 ]
机构
[1] Yonsei Univ, Seoul, South Korea
[2] Hyundai Motor Co, Robot Lab, Seoul, South Korea
关键词
D O I
10.1109/ICCV51070.2023.01052
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We present a novel unsupervised domain adaption method for person re-identification (reID) that generalizes a model trained on a labeled source domain to an unlabeled target domain. We introduce a camera-driven curriculum learning (CaCL) framework that leverages camera labels of person images to transfer knowledge from source to target domains progressively. To this end, we divide target domain dataset into multiple subsets based on the camera labels, and initially train our model with a single subset (i.e., images captured by a single camera). We then gradually exploit more subsets for training, according to a curriculum sequence obtained with a camera-driven scheduling rule. The scheduler considers maximum mean discrepancies (MMD) between each subset and the source domain dataset, such that the subset closer to the source domain is exploited earlier within the curriculum. For each curriculum sequence, we generate pseudo labels of person images in a target domain to train a reID model in a supervised way. We have observed that the pseudo labels are highly biased toward cameras, suggesting that person images obtained from the same camera are likely to have the same pseudo labels, even for different IDs. To address the camera bias problem, we also introduce a camera-diversity (CD) loss encouraging person images of the same pseudo label, but captured across various cameras, to involve more for discriminative feature learning, providing person representations robust to inter-camera variations. Experimental results on standard benchmarks, including real-to-real and synthetic-to-real scenarios, demonstrate the effectiveness of our framework.
引用
收藏
页码:11419 / 11428
页数:10
相关论文
共 50 条
  • [1] Unsupervised domain adaptive person re-identification via camera penalty learning
    Zhu, Xiaodi
    Li, Yanfeng
    Sun, Jia
    Chen, Houjin
    Zhu, Jinlei
    MULTIMEDIA TOOLS AND APPLICATIONS, 2021, 80 (10) : 15215 - 15232
  • [2] Unsupervised domain adaptive person re-identification via camera penalty learning
    Xiaodi Zhu
    Yanfeng Li
    Jia Sun
    Houjin Chen
    Jinlei Zhu
    Multimedia Tools and Applications, 2021, 80 : 15215 - 15232
  • [3] Domain Adaptive Attention Learning for Unsupervised Person Re-Identification
    Huang, Yangru
    Peng, Peixi
    Jin, Yi
    Li, Yidong
    Xing, Junliang
    Ge, Shiming
    THIRTY-FOURTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THE THIRTY-SECOND INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE AND THE TENTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2020, 34 : 11069 - 11076
  • [4] Learning Feature Fusion for Unsupervised Domain Adaptive Person Re-identification
    Ding, Jin
    Zhou, Xue
    2022 26TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR), 2022, : 2613 - 2619
  • [5] Camera Contrast Learning for Unsupervised Person Re-Identification
    Zhang, Guoqing
    Zhang, Hongwei
    Lin, Weisi
    Chandran, Arun Kumar
    Jing, Xuan
    IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY, 2023, 33 (08) : 4096 - 4107
  • [7] Domain-Camera Adaptation for Unsupervised Person Re-Identification
    Tian, Jiajie
    Teng, Zhu
    Li, Yan
    Li, Rui
    Wu, Yi
    Fan, Jianping
    2019 6TH INTERNATIONAL CONFERENCE ON BEHAVIORAL, ECONOMIC AND SOCIO-CULTURAL COMPUTING (BESC 2019), 2019,
  • [8] UNSUPERVISED DOMAIN-ADAPTIVE PERSON RE-IDENTIFICATION WITH MULTI-CAMERA CONSTRAINTS
    Takeuchi, Shun
    Li, Fei
    Iwasaki, Sho
    Ning, Jiaqi
    Suzuki, Genta
    2022 IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING, ICIP, 2022, : 1636 - 1640
  • [9] Representation strategy for unsupervised domain adaptation on person re-identification
    LI Hao
    ZHANG Tao
    LI Shuang
    LI Xuan
    ZHAO Xin
    Optoelectronics Letters, 2024, 20 (12) : 749 - 756
  • [10] Representation strategy for unsupervised domain adaptation on person re-identification
    Hao Li
    Tao Zhang
    Shuang Li
    Xuan Li
    Xin Zhao
    Optoelectronics Letters, 2024, 20 (12) : 749 - 756