3D Semantic Segmentation on ScanNet 3 (val)
57.3mIoUOV3D
Evaluation results
| Method | Links | ||
|---|---|---|---|
| OV3DInput Modality=3D point cloud2025.12 | 57.3 | 72.9 | |
| UNITEInput Modality=RGB2025.12 | 48.7 | 68.3 | |
| OpenScene 3DInput Modality=3D point cloud2025.12 | 46 | 66.3 | |
| PanSt3RInput Modality=RGB, QUBO post-processing=false2025.12 | 42.6 | 49.7 | |
| OpenScene 2DInput Modality=3D point cloud2025.12 | 41.4 | 63.6 | |
| HOV-SGInput Modality=RGB-D2025.12 | 34.4 | 51.1 | |
| LSMInput Modality=RGB2025.12 | 32.2 | 41.1 | |
| CLIP-FO3DInput Modality=3D point cloud2025.12 | 30.2 | 49.1 | |
| Uni3RInput Modality=RGB2025.12 | 29.3 | 39.4 | |
| Concept-GraphsInput Modality=RGB-D2025.12 | 17.1 | 29.1 | |
| Pe3RInput Modality=RGB2025.12 | 10.7 | 19.8 |