Fine-grained Motion Video Understanding on FAVOR-Bench
48.94OverallInternVL3-8B + MotionEnhancer
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| InternVL3-8B + MotionEnhancerFrames=82026.06 | 48.94 | 49.25 | 47.17 | 57.11 | 46.57 | 56.35 | 36.74 | 51.56 | |
| Qwen2.5-VL-72B*Frames=1fps2026.06 | 48.14 | 48.17 | 50.28 | 46.98 | 48.13 | 51.78 | 40.28 | 51.56 | |
| Qwen2.5VL-7B + MotionEnhancerFrames=1fps2026.06 | 46.88 | 47.01 | 49.34 | 50.62 | 45.37 | 53.2 | 30.42 | 53.12 | |
| LLaVA-Video-72B-Qwen2Frames=642026.06 | 46.08 | 46.49 | 48.35 | 47.5 | 45.25 | 51.7 | 33.02 | 53.12 | |
| InternVL3-8B*Frames=82026.06 | 45.82 | 46.35 | 45.39 | 48.54 | 47.59 | 51.45 | 33.58 | 51.56 | |
| Qwen2.5VL-7B + MotionSight*Frames=1fps2026.06 | 45.47 | 45.99 | 46.23 | 51.59 | 45.01 | 50.04 | 29.95 | 53.12 | |
| Qwen2.5VL-3B + MotionEnhancerFrames=1fps2026.06 | 44.53 | 43.94 | 45.01 | 51.59 | 44.4 | 48.96 | 28.37 | 45.31 | |
| VideoChat-Flash-Qwen2-7BFrames=1fps2026.06 | 43.82 | 44.86 | 41.9 | 48.41 | 42.84 | 50.95 | 35.07 | 50 | |
| InternVL3-2B + MotionEnhancerFrames=82026.06 | 43.71 | 45.35 | 38.53 | 54.57 | 42.6 | 51.78 | 33.02 | 51.56 | |
| Qwen2.5VL-7B*Frames=1fps2026.06 | 42.61 | 42.58 | 41.64 | 47.83 | 44.89 | 47.55 | 28.28 | 45.31 | |
| VideoLLaMA3-7BFrames=1fps2026.06 | 41.46 | 41.46 | 40.2 | 44.13 | 42.42 | 48.3 | 31.53 | 42.19 | |
| InternVL3-2B*Frames=82026.06 | 39.27 | 39.11 | 37.66 | 43.28 | 40.49 | 44.98 | 29.21 | 39.06 | |
| LLaVA-Video-7B-Qwen2Frames=642026.06 | 38.6 | 39.94 | 36.14 | 41.27 | 41.28 | 44.48 | 29.58 | 46.88 | |
| Qwen2.5VL-3B*Frames=1fps2026.06 | 37.43 | 38.07 | 38.45 | 38.16 | 39.35 | 43.4 | 23.72 | 45.31 | |
| VideoLLaMA3-2BFrames=1fps2026.06 | 32.98 | 34.61 | 28.97 | 36.6 | 34.9 | 38.01 | 28.56 | 40.62 |