Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2408.17098 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909301371568128 |
|---|---|
| author | Solano-Carrillo, Edgardo Sattler, Felix Alex, Antje Klein, Alexander Costa, Bruno Pereira Rodriguez, Angel Bueno Stoppe, Jannis |
| author_facet | Solano-Carrillo, Edgardo Sattler, Felix Alex, Antje Klein, Alexander Costa, Bruno Pereira Rodriguez, Angel Bueno Stoppe, Jannis |
| contents | The tracking-by-detection paradigm is the mainstream in multi-object tracking, associating tracks to the predictions of an object detector. Although exhibiting uncertainty through a confidence score, these predictions do not capture the entire variability of the inference process. For safety and security critical applications like autonomous driving, surveillance, etc., knowing this predictive uncertainty is essential though. Therefore, we introduce, for the first time, a fast way to obtain the empirical predictive distribution during object detection and incorporate that knowledge in multi-object tracking. Our mechanism can easily be integrated into state-of-the-art trackers, enabling them to fully exploit the uncertainty in the detections. Additionally, novel association methods are introduced that leverage the proposed mechanism. We demonstrate the effectiveness of our contribution on a variety of benchmarks, such as MOT17, MOT20, DanceTrack, and KITTI. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2408_17098 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | UTrack: Multi-Object Tracking with Uncertain Detections Solano-Carrillo, Edgardo Sattler, Felix Alex, Antje Klein, Alexander Costa, Bruno Pereira Rodriguez, Angel Bueno Stoppe, Jannis Computer Vision and Pattern Recognition The tracking-by-detection paradigm is the mainstream in multi-object tracking, associating tracks to the predictions of an object detector. Although exhibiting uncertainty through a confidence score, these predictions do not capture the entire variability of the inference process. For safety and security critical applications like autonomous driving, surveillance, etc., knowing this predictive uncertainty is essential though. Therefore, we introduce, for the first time, a fast way to obtain the empirical predictive distribution during object detection and incorporate that knowledge in multi-object tracking. Our mechanism can easily be integrated into state-of-the-art trackers, enabling them to fully exploit the uncertainty in the detections. Additionally, novel association methods are introduced that leverage the proposed mechanism. We demonstrate the effectiveness of our contribution on a variety of benchmarks, such as MOT17, MOT20, DanceTrack, and KITTI. |
| title | UTrack: Multi-Object Tracking with Uncertain Detections |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2408.17098 |