Synthetic-To-Real Video Person Re-ID
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915135234244608 |
|---|---|
| author | Zhang, Xiangqun Feng, Wei Han, Ruize Wang, Likai Song, Linqi Hou, Junhui |
| author_facet | Zhang, Xiangqun Feng, Wei Han, Ruize Wang, Likai Song, Linqi Hou, Junhui |
| contents | Person re-identification (Re-ID) is an important task and has significant applications for public security and information forensics, which has progressed rapidly with the development of deep learning. In this work, we investigate a novel and challenging setting of Re-ID, i.e., cross-domain video-based person Re-ID. Specifically, we utilize synthetic video datasets as the source domain for training and real-world videos for testing, notably reducing the reliance on expensive real data acquisition and annotation. To harness the potential of synthetic data, we first propose a self-supervised domain-invariant feature learning strategy for both static and dynamic (temporal) features. Additionally, to enhance person identification accuracy in the target domain, we propose a mean-teacher scheme incorporating a self-supervised ID consistency loss. Experimental results across five real datasets validate the rationale behind cross-synthetic-real domain adaptation and demonstrate the efficacy of our method. Notably, the discovery that synthetic data outperforms real data in the cross-domain scenario is a surprising outcome. The code and data are publicly available at https://github.com/XiangqunZhang/UDA_Video_ReID. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_02108 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Synthetic-To-Real Video Person Re-ID Zhang, Xiangqun Feng, Wei Han, Ruize Wang, Likai Song, Linqi Hou, Junhui Computer Vision and Pattern Recognition Person re-identification (Re-ID) is an important task and has significant applications for public security and information forensics, which has progressed rapidly with the development of deep learning. In this work, we investigate a novel and challenging setting of Re-ID, i.e., cross-domain video-based person Re-ID. Specifically, we utilize synthetic video datasets as the source domain for training and real-world videos for testing, notably reducing the reliance on expensive real data acquisition and annotation. To harness the potential of synthetic data, we first propose a self-supervised domain-invariant feature learning strategy for both static and dynamic (temporal) features. Additionally, to enhance person identification accuracy in the target domain, we propose a mean-teacher scheme incorporating a self-supervised ID consistency loss. Experimental results across five real datasets validate the rationale behind cross-synthetic-real domain adaptation and demonstrate the efficacy of our method. Notably, the discovery that synthetic data outperforms real data in the cross-domain scenario is a surprising outcome. The code and data are publicly available at https://github.com/XiangqunZhang/UDA_Video_ReID. |
| title | Synthetic-To-Real Video Person Re-ID |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2402.02108 |