Overall Evaluation on Demo-ICL-Bench
80.1Average ScoreHuman
Evaluation Results
| Method | Links | |
|---|---|---|
| Human2026.02 | 80.1 | |
| Gemini-2.5-Pro2026.02 | 38.9 | |
| GPT-4o2026.02 | 34.9 | |
| Demo-ICLSize=7B, Frame=322026.02 | 33.1 | |
| Demo-ICL (SFT)Size=7B, Frame=322026.02 | 29.8 | |
| Qwen2.5-VLSize=72B, Frame=322026.02 | 29.5 | |
| LLaVA-VideoSize=7B, Frame=322026.02 | 27.2 | |
| VideoChat-R1Size=7B, Frame=322026.02 | 27 | |
| Qwen2.5-VLSize=7B, Frame=322026.02 | 26.3 | |
| Video-R1Size=7B, Frame=322026.02 | 26.2 | |
| InternVL-3Size=8B, Frame=322026.02 | 25 | |
| Ola-Video (Base)Size=7B, Frame=322026.02 | 24.8 | |
| OlaSize=7B, Frame=322026.02 | 24.3 | |
| Qwen2-VLSize=7B, Frame=322026.02 | 22 |