| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| ScanQA (val) | CVP | CIDEr107.1 | 133 | 3d ago | |
| SQA3D (test) | Ross3D | EM@163 | 55 | 3d ago | |
| ScanQA w/ objects (test) | BridgeQA | EM@131.29 | 55 | 3d ago | |
| ScanQA w/o objects (test) | BridgeQA | EM@130.87 | 51 | 3d ago | |
| ScanQA v1.0 (test) | Grounded 3D-LLM | ROUGE72.7 | 26 | 3d ago | |
| ScanQA (test) | NaviLLM | BLEU-413.9 | 20 | 3d ago | |
| NuscenesQA v1.0 (test) | MSMDFusion + MCAN | Existence Accuracy (All)85.4 | 19 | 3d ago | |
| ScanQA | Video-3D LLM (Uniform) | C Score102.1 | 16 | 3d ago | |
| ScanQA v1.0 (val) | Efficient-VLN | BLEU-416.2 | 13 | 3d ago | |
| SQA3D | Video-3D LLM (Uniform) | EM58.6 | 11 | 3d ago | |
| ObjaXL-LVIS 1K (test) | GPT4Point (FLANT5XL) | Accuracy27.6 | 8 | 3d ago | |
| SQA3D v1.0 (test) | PAVE-7B (w/ 3D info) | EM@159 | 8 | 3d ago | |
| MMScan QA | LLaVA-3D | Overall Score55.4 | 7 | 3d ago | |
| ScanQA Leaderboard as of Nov 2023 (test) | EM30.82 | 5 | 3d ago | ||
| 3DQA (val) | DSPNet | Exact Match @ 152 | 2 | 3d ago | |
| Cap3D QA | Ours | Accuracy39.3 | 1 | 3d ago |
