Video Object Discovery and Tracking on OVIS
56.3ARIfgTSA
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| TSAK=22, Frames=Up to 500, Backbone=DINOv2 ViT-S/14, Resolution=256x2562026.06 | 56.3 | 30.7 | 21.6 | 19 | |
| RandSF.QsscK=22, Frames=Up to 500, Backbone=DINOv2 ViT-S/14, Resolution=256x2562026.06 | 30.4 | 18.6 | 7.6 | 3 | |
| SlotContrastK=22, Frames=Up to 500, Backbone=DINOv2 ViT-S/14, Resolution=256x2562026.06 | 24.3 | 16.1 | 6.5 | 1.5 | |
| VideoSAURK=22, Frames=Up to 500, Backbone=DINOv2 ViT-S/14, Resolution=256x2562026.06 | 23.4 | 14.1 | 5.8 | 1.4 | |
| RandSF.QtsimK=22, Frames=Up to 500, Backbone=DINOv2 ViT-S/14, Resolution=256x2562026.06 | 22.5 | 16.2 | 8.1 | 4.3 |