Fig 1.
HOTA and IDF1 metric results on MOT17 (left) and MOT20 (right) test sets.
Fig 2.
Overview of the proposed model architecture.
Fig 3.
Each row shows one sequence: the object before occlusion, during occlusion, and after reappearing with the same ID.
Table 1.
Ablation results for individual components on the MOT17 public validation split.
Table 2.
Individual Component Ablation on MOT20 (Validation Split).
Fig 4.
Ablation study: individual component performance on the MOT17 validation set.
Fig 5.
Ablation study: individual component performance on the MOT20 validation set.
Table 3.
Cumulative component ablation on MOT17 (Public Validation Split).
Fig 6.
Ablation study: cumulative component performance on the MOT17 validation set.
Table 4.
Cumulative component ablation on MOT20 (Validation Split).
Fig 7.
Ablation study: cumulative component performance on the MOT20 validation set.
Table 5.
Comparison of state-of-the-art methods on MOT17 and MOT20 benchmarks.