Video Reasoning Segmentation on VideoReasonSeg Short (val)
62.6Seg J&FViLLa
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ViLLaBackbone=InternVideo2-6B2024.07 | 62.6 | 53.8 | |
| VISABackbone=InternVideo2-6B2024.07 | 57.9 | 44.6 | |
| ViLLaBackbone=InternVideo2-1B2024.07 | 57.4 | 46.5 | |
| VideoLISA*Backbone=InternVideo2-6B2024.07 | 54.8 | 43.6 | |
| VISABackbone=Chat-UniVi-13B2024.07 | 54.8 | 42.6 | |
| VISABackbone=LLaVA-13B2024.07 | 54.6 | 40.1 | |
| VideoLISA*Backbone=InternVideo2-1B2024.07 | 51 | 38.4 | |
| VISABackbone=LLaVA-7B2024.07 | 48.9 | 37.5 | |
| LISABackbone=LLaVA-13B2024.07 | 45 | 35.2 | |
| LISABackbone=LLaVA-7B2024.07 | 43.8 | 34.5 | |
| OnlineRefer2024.07 | 26.9 | — | |
| LMPM2024.07 | 26.4 | — | |
| ReferFormer2024.07 | 17.5 | — | |
| MTTR2024.07 | 17.2 | — |