Unsupervised Video Object Discovery on YTVIS-HQ unconditional (test)
45.5ARISmoothSA+xSSC
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| SmoothSA+xSSCEncoder=DINOv2 ViT-S/14, Input Resolution=224x224, #slot=6, Conditionality=unconditional2026.05 | 45.5 | 61.4 | 41.1 | 40.4 | |
| SmoothSAEncoder=DINOv2 ViT-S/14, Input Resolution=224x224, #slot=6, Conditionality=unconditional2026.05 | 42.4 | 63 | 38.9 | 38.3 | |
| RandSF.Q+xSSCEncoder=DINOv2 ViT-S/14, Input Resolution=224x224, #slot=6, Conditionality=unconditional2026.05 | 41.8 | 58.2 | 39.5 | 39.2 | |
| RandSF.QEncoder=DINOv2 ViT-S/14, Input Resolution=224x224, #slot=6, Conditionality=unconditional2026.05 | 40.1 | 58 | 37.6 | 37.2 | |
| SlotContrast+xSSCEncoder=DINOv2 ViT-S/14, Input Resolution=224x224, #slot=6, Conditionality=unconditional2026.05 | 39.3 | 52.5 | 35.1 | 34.3 | |
| SlotContrastEncoder=DINOv2 ViT-S/14, Input Resolution=224x224, #slot=6, Conditionality=unconditional2026.05 | 37.2 | 49.4 | 33 | 32.8 | |
| VideoSAUREncoder=DINOv2 ViT-S/14, Input Resolution=224x224, #slot=6, Conditionality=unconditional2026.05 | 33.8 | 49.2 | 29.9 | 29.7 |