Video Reasoning on Perception Test
67.2AccuracyBoxTuning
Evaluation Results
| Method | Links | |
|---|---|---|
| BoxTuningSize=7B2026.04 | 67.2 | |
| ObjectMLLMSize=7B2026.04 | 66.6 | |
| SeViLASize=3B2026.04 | 62 | |
| VideoLLaMA2Size=7B, Zero-shot=true2026.04 | 51.4 | |
| LLaVA-NeXT-Vid.Size=7B, Zero-shot=true2026.04 | 49.3 |