3D/4D Video Question Answering on SAT
64.8AccuracyGemini-1.5-Pro
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini-1.5-Pro2025.12 | 64.8 | |
| 4D-RGPT-8Bparameters=8B2025.12 | 64.7 | |
| LLaVA-Video-7B + SATparameters=7B, enhancement=SAT2025.12 | 63.4 | |
| NVILA-Lite-8Bparameters=8B2025.12 | 62 | |
| GPT-4o2025.12 | 57.5 | |
| LLaVA-Video-7Bparameters=7B2025.12 | 53.5 | |
| SpaceR-7Bparameters=7B2025.12 | 47.8 | |
| LLaVA-OneVision-7Bparameters=7B2025.12 | 41.7 |