Video Understanding on VDC (MAT and Wall-time speedup)
8.05MATHIPPO
Evaluation results
| Method | Links | ||
|---|---|---|---|
| HIPPOBackbone Model=LLaVA-OneVision 7&72B2026.01 | 8.05 | 3.31 | |
| HIPPOBackbone Model=Video-SALMONN2+ 7&72B2026.01 | 6.97 | 2.31 | |
| HIPPOBackbone Model=Qwen2.5-VL 7&72B2026.01 | 6.29 | 2.31 | |
| HIPPOBackbone Model=Qwen3-VL 7&32B2026.01 | 4.91 | 2.35 | |
| SD-treeBackbone Model=Qwen3-VL 7&32B2026.01 | 4.25 | 1.63 | |
| SpecVLMBackbone Model=Qwen3-VL 7&32B2026.01 | 4.01 | 2.05 | |
| Vanilla SDBackbone Model=Qwen3-VL 7&32B2026.01 | 3.79 | 1.27 | |
| SD-treeBackbone Model=LLaVA-OneVision 7&72B2026.01 | 3.68 | 2.08 | |
| SD-treeBackbone Model=Video-SALMONN2+ 7&72B2026.01 | 3.64 | 1.73 | |
| SD-treeBackbone Model=Qwen2.5-VL 7&72B2026.01 | 3.64 | 1.72 | |
| SpecVLMBackbone Model=LLaVA-OneVision 7&72B2026.01 | 3.48 | 2.68 | |
| SpecVLMBackbone Model=Video-SALMONN2+ 7&72B2026.01 | 3.42 | 2.01 | |
| SpecVLMBackbone Model=Qwen2.5-VL 7&72B2026.01 | 3.42 | 2.11 | |
| Vanilla SDBackbone Model=Qwen2.5-VL 7&72B2026.01 | 3.37 | 1.56 | |
| Vanilla SDBackbone Model=LLaVA-OneVision 7&72B2026.01 | 3.32 | 1.89 | |
| Vanilla SDBackbone Model=Video-SALMONN2+ 7&72B2026.01 | 2.76 | 1.41 |