Fine-grained Video Understanding on FineVidBench (test)
73.1ActionVideo-CCAM
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Video-CCAMParams.=9B2025.04 | 73.1 | 55.9 | 31.65 | 23.45 | 45.66 | 64.95 | 90.27 | 22.72 | 53.55 | 48.47 | 50.96 | |
| Qwen2-VLParams.=7B2025.04 | 68.18 | 57.14 | 24.62 | 96.65 | 33.33 | 74.53 | 90.7 | 22.48 | 49.98 | 63.54 | 58.45 | |
| Video-LLaMA 2.1Params.=7B2025.04 | 63.26 | 50.92 | 19.89 | 30.17 | 42.27 | 76.01 | 89.92 | 26.87 | 44.69 | 53.05 | 49.91 | |
| ShareGPT4VideoParams.=8B2025.04 | 46.9 | 43.88 | 31.76 | 26.33 | 61.05 | 88.44 | 84.8 | 23.36 | 40.85 | 57.11 | 50.82 | |
| MiniCPM-V 2.6Params.=8B2025.04 | 43.37 | 52.56 | 19.13 | 90.32 | 56.42 | 75.66 | 76.49 | 18.09 | 38.35 | 63.4 | 54.01 | |
| LLaVA-NeXT-VideoParams.=7B2025.04 | 37.31 | 42.67 | 22.35 | 19.37 | 24.02 | 53.75 | 75.45 | 20.67 | 34.11 | 38.65 | 36.95 | |
| (Random)2025.04 | 25 | 25 | 25 | 25 | 25 | 33.33 | 33.33 | 25 | 25 | 28.33 | 27.08 |