Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2023
|
| Schlagworte: | |
| Online-Zugang: | https://arxiv.org/abs/2307.14591 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910904575066112 |
|---|---|
| author | Huang, Junchao Wu, Xiaoqi He Yebo Zhao, Sheng |
| author_facet | Huang, Junchao Wu, Xiaoqi He Yebo Zhao, Sheng |
| contents | The purpose of multi-object tracking (MOT) is to continuously track and identify objects detected in videos. Currently, most methods for multi-object tracking model the motion information and combine it with appearance information to determine and track objects. In this paper, unfalsified control is employed to address the ID-switch problem in multi-object tracking. We establish sequences of appearance information variations for the trajectories during the tracking process and design a detection and rectification module specifically for ID-switch detection and recovery. We also propose a simple and effective strategy to address the issue of ambiguous matching of appearance information during the data association process. Experimental results on publicly available MOT datasets demonstrate that the tracker exhibits excellent effectiveness and robustness in handling tracking errors caused by occlusions and rapid movements. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2307_14591 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | The detection and rectification for identity-switch based on unfalsified control Huang, Junchao Wu, Xiaoqi He Yebo Zhao, Sheng Computer Vision and Pattern Recognition Artificial Intelligence The purpose of multi-object tracking (MOT) is to continuously track and identify objects detected in videos. Currently, most methods for multi-object tracking model the motion information and combine it with appearance information to determine and track objects. In this paper, unfalsified control is employed to address the ID-switch problem in multi-object tracking. We establish sequences of appearance information variations for the trajectories during the tracking process and design a detection and rectification module specifically for ID-switch detection and recovery. We also propose a simple and effective strategy to address the issue of ambiguous matching of appearance information during the data association process. Experimental results on publicly available MOT datasets demonstrate that the tracker exhibits excellent effectiveness and robustness in handling tracking errors caused by occlusions and rapid movements. |
| title | The detection and rectification for identity-switch based on unfalsified control |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2307.14591 |