Video Reasoning Segmentation on ReVOS
57J&F ScoreViLLa-InternVideo2-6B
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ViLLa-InternVideo2-6BBackbone=InternVideo2-6B2024.07 | 57 | 54.9 | 59.1 | |
| VISA-InternVideo2-6BBackbone=InternVideo2-6B2024.07 | 52.4 | 50.1 | 54.7 | |
| VISA-Chat-UniVi-13BBackbone=Chat-UniVi-13B2024.07 | 50.9 | 48.8 | 52.9 | |
| LISA-LLaVA-13BBackbone=LLaVA-13B2024.07 | 41.8 | 39.6 | 43.9 | |
| EVISBackbone=InternVL2-1B2026.06 | 40.3 | 38 | 42.6 | |
| VideoLISABackbone=LLaVA-Phi-3-3.8B2026.06 | 39.5 | 37.3 | 41.7 | |
| VISABackbone=Chat-UniVi-7B2026.06 | 39.2 | 36.7 | 41.7 | |
| TrackGPTBackbone=LLaVA-7B2026.06 | 39 | 36.8 | 41.2 | |
| LISABackbone=LLaVA-7B2026.06 | 36.1 | 33.8 | 38.4 | |
| ReferFormerBackbone=Video-Swin-B2026.06 | 23.4 | 21.3 | 25.6 | |
| MTTRBackbone=Video-Swin-T2026.06 | 21 | 20.4 | 21.5 | |
| LMPMBackbone=Swin-T2026.06 | 18.8 | 13.3 | 24.3 |