Diffusion Augmentation and Pose Generation Based Pre-Training Method for Robust Visible-Infrared Person Re-Identification

被引:2
作者
Sun, Rui [1 ]
Huang, Guoxi [2 ]
Xie, Ruirui [2 ]
Wang, Xuebin [2 ]
Chen, Long [2 ]
机构
[1] Hefei Univ Technol, Sch Comp & Informat, Anhui Prov Key Lab Ind Safety & Emergency Technol, Key Lab Knowledge Engn Big Data,Minist Educ, Hefei 230009, Peoples R China
[2] Hefei Univ Technol, Sch Comp & Informat, Anhui Prov Key Lab Ind Safety & Emergency Technol, Hefei 230009, Anhui, Peoples R China
基金
中国国家自然科学基金;
关键词
Person re-identification; visible-infrared; self-supervised; corruption robustness; pre-; training;
D O I
10.1109/LSP.2024.3466792
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
Cross-Modal Visible-Infrared Person Re-identification (VI-REID) constitutes a vital application for constructing all-time surveillance systems. However, the current VI-REID model exhibits significant performance deterioration in noisy environments. Existing algorithms endeavor to mitigate this challenge through fine-tuning stages. We contend that, in contrast to fine-tuning stages, the pre-training phase can effectively exploit the attributes of extensive unlabeled data, thereby facilitating the development of a robust VI-REID model. Therefore, in this paper, we propose a pre-training method for VI-REID based on Diffusion Augmentation and Pose Generation (DAPG), aiming to enhance the robustness and recognition rate of VI-REID models in the presence of damaged scenes. Multiple transfer experiments on the SYSU-MM01 and RegDB datasets demonstrate that our method outperforms existing self-supervised methods, as evidenced by the results.
引用
收藏
页码:2670 / 2674
页数:5
相关论文
共 50 条
[41]   Learn Robust Pedestrian Representation Within Minimal Modality Discrepancy for Visible-Infrared Person Re-Identification [J].
Yu-Jie Liu ;
Wen-Bin Shao ;
Xiao-Rui Sun .
Journal of Computer Science and Technology, 2022, 37 :641-651
[42]   Dual-Path Part-Level Method for Visible-Infrared Person Re-identification [J].
Xiang, Xuezhi ;
Lv, Ning ;
Zhai, Mingliang ;
Abdeen, Rokia ;
El Saddik, Abdulmotaleb .
NEURAL PROCESSING LETTERS, 2020, 52 (01) :313-328
[43]   Learn Robust Pedestrian Representation Within Minimal Modality Discrepancy for Visible-Infrared Person Re-Identification [J].
Liu, Yu-Jie ;
Shao, Wen-Bin ;
Sun, Xiao-Rui .
JOURNAL OF COMPUTER SCIENCE AND TECHNOLOGY, 2022, 37 (03) :641-651
[44]   Augmented Deep Multi-Granularity Pose-Aware Feature Fusion Network for Visible-Infrared Person Re-Identification [J].
Shi, Zheng ;
Song, Wanru ;
Shan, Junhao ;
Liu, Feng .
CMC-COMPUTERS MATERIALS & CONTINUA, 2023, 77 (03) :3467-3488
[45]   Combining information augmentation aggregation and dual-granularity feature fusion for visible-infrared person re-identification [J].
Tao, Mengzhu ;
Long, Huiyun ;
Kong, Guangqian ;
Duan, Xun .
SIGNAL IMAGE AND VIDEO PROCESSING, 2025, 19 (01)
[46]   Joint Modal Alignment and Feature Enhancement for Visible-Infrared Person Re-Identification [J].
Lin, Ronghui ;
Wang, Rong ;
Zhang, Wenjing ;
Wu, Ao ;
Bi, Yihan .
SENSORS, 2023, 23 (11)
[47]   Identity Consistency Construction for Visible-Infrared Person Re-identification in Cloud Environment [J].
Wang, Yiming ;
Xu, Kaixiong ;
Chai, Yi ;
Li, Shuo ;
Jiang, Yutao ;
Liu, Bowen .
PROCEEDINGS OF 2023 CHINESE INTELLIGENT SYSTEMS CONFERENCE, VOL III, 2023, :799-807
[48]   Cross-modality consistency learning for visible-infrared person re-identification [J].
Shao, Jie ;
Tang, Lei .
JOURNAL OF ELECTRONIC IMAGING, 2022, 31 (06)
[49]   Learning dual attention enhancement feature for visible-infrared person re-identification [J].
Zhang, Guoqing ;
Zhang, Yinyin ;
Zhang, Hongwei ;
Chen, Yuhao ;
Zheng, Yuhui .
JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION, 2024, 99
[50]   Visible-Infrared Cross-Modality Person Re-Identification via Adaptive Weighted Triplet Loss and Progressive Training [J].
Song, Ling ;
Yu, Minggong ;
Sun, Delin ;
Zhong, Xionghu .
IEEE ACCESS, 2024, 12 :181799-181807