3D Semantic Segmentation on ScanNet B15 N4
73.3hIoUFully-Sup.
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Fully-Sup.CN prior=true2022.11 | 73.3 | 68.4 | 79.1 | |
| Fully-Sup.C^N prior=Yes2023.08 | 73.3 | 68.4 | 79.1 | |
| PLA (w/ self-train)CN prior=true2022.11 | 70.3 | 68.9 | 71.7 | |
| PLACN prior=false2022.11 | 65.3 | 68.3 | 62.4 | |
| Lowis3DC^N prior=No2023.08 | 65.3 | 68.3 | 62.4 | |
| PLA (w/o Cap.)CN prior=false2022.11 | 39.7 | 68.3 | 28 | |
| PLA (w/o Cap.)C^N prior=No2023.08 | 39.7 | 68.3 | 28 | |
| 3DGenZCN prior=true2022.11 | 20.6 | 56 | 12.6 | |
| 3DGenZC^N prior=Yes2023.08 | 20.6 | 56 | 12.6 | |
| 3DTZSLCN prior=true2022.11 | 10.5 | 36.7 | 6.1 | |
| 3DTZSLC^N prior=Yes2023.08 | 10.5 | 36.7 | 6.1 | |
| LSeg-3DCN prior=false2022.11 | 0 | 64.4 | 0 | |
| OV-SparseConvNetC^N prior=No2023.08 | 0 | 64.4 | 0 |