Progressive Discriminative Feature Learning for Visible-Infrared Person Re-Identification

被引:0
|
作者
Zhou, Feng [1 ]
Cheng, Zhuxuan [2 ]
Yang, Haitao [1 ]
Song, Yifeng [2 ]
Fu, Shengpeng [2 ]
机构
[1] Hunan Police Acad, Dept Criminal Ivest, Changsha 410138, Peoples R China
[2] Chinese Acad Sci, Shenyang Inst Automat, Shenyang 110016, Peoples R China
基金
中国国家自然科学基金;
关键词
visible-infrared person re-identification; cross-modality; deep learning; instance normalization;
D O I
10.3390/electronics13142825
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The visible-infrared person re-identification (VI-ReID) task aims to retrieve the same pedestrian between visible and infrared images. VI-ReID is a challenging task due to the huge modality discrepancy and complex intra-modality variations. Existing works mainly complete the modality alignment at one stage. However, aligning modalities at different stages has positive effects on the intra-class and inter-class distances of cross-modality features, which are often ignored. Moreover, discriminative features with identity information may be corrupted in the processing of modality alignment, further degrading the performance of person re-identification. In this paper, we propose a progressive discriminative feature learning (PDFL) network that adopts different alignment strategies at different stages to alleviate the discrepancy and learn discriminative features progressively. Specifically, we first design an adaptive cross fusion module (ACFM) to learn the identity-relevant features via modality alignment with channel-level attention. For well preserving identity information, we propose a dual-attention-guided instance normalization module (DINM), which can well guide instance normalization to align two modalities into a unified feature space through channel and spatial information embedding. Finally, we generate multiple part features of a person to mine subtle differences. Multi-loss optimization is imposed during the training process for more effective learning supervision. Extensive experiments on the public datasets of SYSU-MM01 and RegDB validate that our proposed method performs favorably against most state-of-the-art methods.
引用
收藏
页数:15
相关论文
共 50 条
  • [21] Progressive Cross-Modal Association Learning for Unsupervised Visible-Infrared Person Re-Identification
    Yang, Yiming
    Hu, Weipeng
    Hu, Haifeng
    IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, 2025, 20 : 1290 - 1304
  • [22] Hybrid Modality Metric Learning for Visible-Infrared Person Re-Identification
    Zhang, La
    Guo, Haiyun
    Zhu, Kuan
    Qiao, Honglin
    Huang, Gaopan
    Zhang, Sen
    Zhang, Huichen
    Sun, Jian
    Wang, Jinqiao
    ACM TRANSACTIONS ON MULTIMEDIA COMPUTING COMMUNICATIONS AND APPLICATIONS, 2022, 18 (01)
  • [23] Robust Duality Learning for Unsupervised Visible-Infrared Person Re-Identification
    Li, Yongxiang
    Sun, Yuan
    Qin, Yang
    Peng, Dezhong
    Peng, Xi
    Hu, Peng
    IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, 2025, 20 : 1937 - 1948
  • [24] Contrastive Learning with Information Compensation for Visible-Infrared Person Re-Identification
    Zhang, La
    Guo, Haiyun
    Zhao, Xu
    Sun, Jian
    Wang, Jinqiao
    2024 14TH ASIAN CONTROL CONFERENCE, ASCC 2024, 2024, : 1266 - 1271
  • [25] Four-Stream Network and Nonsignificant Feature Learning for Visible-Infrared Person Re-Identification
    Liang, Yilei
    Han, Hua
    Huang, Li
    Wang, Chunyuan
    INTERNATIONAL JOURNAL OF PATTERN RECOGNITION AND ARTIFICIAL INTELLIGENCE, 2022, 36 (07)
  • [26] Fine-grained Learning for Visible-Infrared Person Re-identification
    Qi, Mengzan
    Chan, Sixian
    Hang, Chen
    Zhang, Guixu
    Li, Zhi
    2023 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO, ICME, 2023, : 2417 - 2422
  • [27] Learning with Twin Noisy Labels for Visible-Infrared Person Re-Identification
    Yang, Mouxing
    Huang, Zhenyu
    Hu, Peng
    Li, Taihao
    Lv, Jiancheng
    Peng, Xi
    2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2022, : 14288 - 14297
  • [28] Attributes Based Visible-Infrared Person Re-identification
    Zheng, Aihua
    Feng, Mengya
    Pan, Peng
    Jiang, Bo
    Luo, Bin
    PATTERN RECOGNITION AND COMPUTER VISION, PT I, PRCV 2022, 2022, 13534 : 254 - 266
  • [29] Minimizing Maximum Feature Space Deviation for Visible-Infrared Person Re-Identification
    Wu, Zhixiong
    Wen, Tingxi
    APPLIED SCIENCES-BASEL, 2022, 12 (17):
  • [30] Joint Modal Alignment and Feature Enhancement for Visible-Infrared Person Re-Identification
    Lin, Ronghui
    Wang, Rong
    Zhang, Wenjing
    Wu, Ao
    Bi, Yihan
    SENSORS, 2023, 23 (11)