3D Question Answering on ScanQA (evaluation)
107.4CIDErV-LynX-7B
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| V-LynX-7BEvaluation Protocol=Task-specific models >= 7B, Delta Params=195.0M2026.05 | 107.4 | 16.7 | 20.8 | 50.3 | 29.7 | 48.6 | |
| PAVE-7BEvaluation Protocol=Task-specific models >= 7B, Delta Params=475.0M2026.05 | 103.4 | 16 | 19.9 | 49 | 29.1 | 48.5 | |
| LEO-7BEvaluation Protocol=Task-specific models >= 7B, Delta Params=-2026.05 | 101.4 | 13.2 | 20 | 49.2 | 24.5 | 47.6 | |
| LLaVA-OV-7B-FTEvaluation Protocol=Task-specific models >= 7B, Delta Params=161.5M2026.05 | 95.1 | 13.5 | 19.1 | 47.4 | 27.4 | 46.3 | |
| LLaVA-3D-7BEvaluation Protocol=Task-specific models >= 7B, Delta Params=-2026.05 | 91.7 | 14.5 | 20.7 | 50.1 | 27 | 45 | |
| LLaVA-OV-7BEvaluation Protocol=Zero-shot Video LLMs, Delta Params=-2026.05 | 91 | 5.3 | 18.2 | 45.9 | 26.7 | — | |
| V-LynX-0.5BEvaluation Protocol=Task-specific models < 7B, Delta Params=68.7M2026.05 | 87.1 | 14.3 | 17.2 | 43.8 | 26.4 | 44.2 | |
| PAVE-0.5BEvaluation Protocol=Task-specific models < 7B, Delta Params=345.9M2026.05 | 84.2 | 13.1 | 17 | 42.1 | 23.1 | 40 | |
| Scene-LLM-7BEvaluation Protocol=Task-specific models >= 7B, Delta Params=-2026.05 | 80 | 12 | 16.6 | 40 | 27.2 | — | |
| 3D-LLM-7BEvaluation Protocol=Task-specific models >= 7B, Delta Params=-2026.05 | 74.5 | 12.9 | 15.1 | 37.5 | 21.2 | — | |
| LLaVA-OV-0.5B-FTEvaluation Protocol=Task-specific models < 7B, Delta Params=35.2M2026.05 | 70.5 | 6.5 | 14.3 | 36.9 | 20.1 | 36.3 | |
| VideoChat2-7BEvaluation Protocol=Zero-shot Video LLMs, Delta Params=-2026.05 | 49.2 | 9.6 | 9.5 | 28.2 | 19.2 | — | |
| LLaVA-OV-0.5BEvaluation Protocol=Zero-shot Video LLMs, Delta Params=-2026.05 | 17.2 | 1.2 | 13.7 | 18.4 | 0.2 | 28 |