Situated 3D Question Answering on SQA3D
62.3EMCVP
Evaluation results
| Method | Links | ||
|---|---|---|---|
| CVP2025.12 | 62.3 | — | |
| 3DGraphLLM-7B + QuatRoPEDetector / Segmentation=Mask3D2026.03 | 55.2 | — | |
| Chat-Scene-7B + QuatRoPEDetector / Segmentation=Mask3D2026.03 | 54.7 | — | |
| Chat-SceneModel Category=LLM-based Models2023.12 | 54.6 | 57.5 | |
| Chat-Scene-7BDetector / Segmentation=Mask3D2026.03 | 54.6 | — | |
| Scene-LLMModel Category=LLM-based Models2023.12 | 54.2 | — | |
| Scene-LLMDetector / Segmentation=N/A2026.03 | 53.6 | — | |
| 3DGraphLLM-1B + QuatRoPEDetector / Segmentation=GT2026.03 | 53.2 | — | |
| Chat-Scene-1B + QuatRoPEDetector / Segmentation=GT2026.03 | 53.1 | — | |
| 3DGraphLLM-7BDetector / Segmentation=Mask3D2026.03 | 53.1 | — | |
| BridgeQADetector / Segmentation=VoteNet2026.03 | 52.9 | — | |
| Chat-Scene++Input Modality=2D-only2026.03 | 52.9 | — | |
| 3DGraphLLM-1BDetector / Segmentation=GT2026.03 | 51.1 | — | |
| Chat-Scene-1BDetector / Segmentation=GT2026.03 | 50.7 | — | |
| 3D-VisTAModel Category=Expert Models2023.12 | 48.5 | — | |
| 3D-VisTADetector / Segmentation=Mask3D2026.03 | 48.5 | — | |
| PQ3DDetector / Segmentation=PQ3D Promptable2026.03 | 47.1 | — | |
| GPT-4otask-specific fine-tuning=none2025.12 | 41.7 | — | |
| LEOModel Category=LLM-based Models2023.12 | — | 53.7 |