Visual Question Answering (VQA) on 6-DoF SpatialBench
Metric: Total (higher is better)
Metric: Total (higher is better)
| # | Model↕ | Total▼ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | SoFar | 43.9 | No | SoFar: Language-Grounded Orientation Bridges Spa... | 2025-02-18 | Code |
| 2 | GPT-4o | 36.2 | No | GPT-4o System Card | 2024-10-25 | - |
| 3 | RoboPoint | 33.5 | No | RoboPoint: A Vision-Language Model for Spatial A... | 2024-06-15 | - |
| 4 | SpatialBot | 32.7 | No | SpatialBot: Precise Spatial Understanding with V... | 2024-06-19 | Code |
| 5 | SpaceMantis | 28.9 | No | SpatialVLM: Endowing Vision-Language Models with... | 2024-01-22 | - |
| 6 | SpaceLLaVA | 28.2 | No | SpatialVLM: Endowing Vision-Language Models with... | 2024-01-22 | - |
| 7 | LLaVA-1.5 | 27.2 | No | Improved Baselines with Visual Instruction Tuning | 2023-10-05 | Code |