| Benchmark | Model | Rank | Results |
|---|---|---|---|
| video-instance-segmentation-on-youtube-vis-2021 | LVMT ViT-L (DINOv2) | #1 | mask AP: 66.4 |
| video-panoptic-segmentation-on-vipseg | LVMT ViT-L (DINOv3) | #1 | VPQ: 60.3 |
| video-semantic-segmentation-on-vspw | LVMT ViT-L (DINOv3) | #1 | mIoU: 66.4 |
| video-instance-segmentation-on-youtube-vis-validation | LVMT ViT-L (DINOv2) | #2 | mask AP: 70.1 |
| video-instance-segmentation-on-ovis-validation | LVMT ViT-L (DINOv3) | #3 | mask AP: 56.7 |