Open-vocabulary 3D Instance Segmentation on ScanNet200 (val)
26.9mAPMask3D (supervised)
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| Mask3D (supervised)3D proposals (Point-based)=true2025.12 | 26.9 | 36.2 | 41.4 | 39.8 | 21.7 | 17.9 | 13.41 | |
| Any3DIS3D proposals (Point-based)=true, 3D proposals (RGBD-based)=true, 2D detector=true, SAM (Proposal generation)=true, CLIP=true2025.12 | 25.8 | — | — | 27.4 | 23.8 | 26.4 | — | |
| BoxOVIS3D proposals (Point-based)=true, 3D proposals (RGBD-based)=true, 2D detector=true2025.12 | 24.9 | 32.1 | 36.8 | 27.6 | 24.3 | 22.4 | 55.9 | |
| Open-YOLO3D3D proposals (Point-based)=true, 2D detector=true2025.12 | 24.7 | 31.7 | 36.2 | 27.8 | 24.3 | 21.6 | 21.8 | |
| Open3DIS3D proposals (Point-based)=true, 3D proposals (RGBD-based)=true, 2D detector=true, SAM (Proposal generation)=true, CLIP=true2025.12 | 23.7 | 29.4 | 32.8 | 27.8 | 21.2 | 21.8 | 360.12 | |
| OpenMask3D3D proposals (Point-based)=true, SAM (Mask labeling)=true, CLIP=true2025.12 | 15.4 | 19.9 | 23.1 | 17.1 | 14.1 | 14.9 | 553.87 | |
| OVIR-3D3D proposals (RGBD-based)=true, 2D detector=true, SAM (Proposal generation)=true, CLIP=true2025.12 | 13 | 24.9 | 32.3 | 14.4 | 12.7 | 11.7 | 466.8 |