Nir Aharon, Roy Orfaig, Ben-Zion Bobrovsky
The goal of multi-object tracking (MOT) is detecting and tracking all the objects in a scene, while keeping a unique identifier for each object. In this paper, we present a new robust state-of-the-art tracker, which can combine the advantages of motion and appearance information, along with camera-motion compensation, and a more accurate Kalman filter state vector. Our new trackers BoT-SORT, and BoT-SORT-ReID rank first in the datasets of MOTChallenge [29, 11] on both MOT17 and MOT20 test sets, in terms of all the main MOT metrics: MOTA, IDF1, and HOTA. For MOT17: 80.5 MOTA, 80.2 IDF1, and 65.0 HOTA are achieved. The source code and the pre-trained models are available at https://github.com/NirAharon/BOT-SORT
| Task | Dataset | Metric | Value | Model |
|---|---|---|---|---|
| Multi-Object Tracking | MOT20 | HOTA | 63.3 | BoT-SORT |
| Multi-Object Tracking | MOT20 | IDF1 | 77.5 | BoT-SORT |
| Multi-Object Tracking | MOT20 | MOTA | 77.8 | BoT-SORT |
| Multi-Object Tracking | MOT17 | HOTA | 65 | BoT-SORT |
| Multi-Object Tracking | MOT17 | IDF1 | 80.2 | BoT-SORT |
| Multi-Object Tracking | MOT17 | MOTA | 80.5 | BoT-SORT |
| Object Tracking | QuadTrack | HOTA | 15.77 | Bot-SORT |
| Object Tracking | MOT20 | HOTA | 63.3 | BoT-SORT |
| Object Tracking | MOT20 | IDF1 | 77.5 | BoT-SORT |
| Object Tracking | MOT20 | MOTA | 77.8 | BoT-SORT |
| Object Tracking | MOT17 | HOTA | 65 | BoT-SORT |
| Object Tracking | MOT17 | IDF1 | 80.2 | BoT-SORT |
| Object Tracking | MOT17 | MOTA | 80.5 | BoT-SORT |