Instance Segmentation on ScanNet200 v2 (val)
31.5mAPODIN
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| ODINInput Data Type=Sensor RGBD Point Cloud, Backbone=Swin-B2024.01 | 31.5 | 45.3 | 53.1 | |
| MAFTInput Data Type=Mesh Sampled Point Cloud2024.01 | 29.2 | 38.2 | 43.3 | |
| QueryFormerInput Data Type=Mesh Sampled Point Cloud2024.01 | 28.1 | 37.1 | 43.4 | |
| Mask3DInput Data Type=Mesh Sampled Point Cloud2024.01 | 27.4 | 37 | 42.3 | |
| ODINInput Data Type=Sensor RGBD Point Cloud, Backbone=ResNet502024.01 | 25.6 | 36.9 | 43.8 | |
| Mask3DInput Data Type=Sensor RGBD Point Cloud, Trained by this paper=true2024.01 | 15.5 | 21.4 | 24.3 | |
| OpenMask3DInput Data Type=Zero-Shot2024.01 | 15.4 | 19.9 | 23.1 |