3D Semantic Segmentation on Matterport3D (val)
54.12mIoUMinkowskiNet (reimple)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MinkowskiNet (reimple)Type=Upper-bound2025.03 | 54.12 | 65.57 | |
| PTv3-Matterport3DParams=46M, Activated=46M, Training Setting=Single-dataset Training, Precise Evaluator=false2025.05 | 53.6 | — | |
| SupConType=Fully-sup.2025.03 | 53.1 | 63.4 | |
| MinkowskiNetType=Fully-sup.2025.03 | 53.1 | 63.4 | |
| OpenScene-2D3DType=Zero-shot, variant=original paper based on Lseg2025.03 | 43.4 | 53.5 | |
| CUA-O3D (2D3D)Type=Zero-shot2025.03 | 42.2 | 50.9 | |
| OpenScene-3DType=Zero-shot, variant=original paper based on Lseg2025.03 | 41.9 | 51.2 | |
| OpenScene (reimple)-2D3DType=Zero-shot2025.03 | 41.5 | 50.6 | |
| CUA-O3D (3D)Type=Zero-shot2025.03 | 41.3 | 49.5 | |
| Point-MoE-LTraining Setup=Indoor and Outdoor joint training, Evaluation Protocol=Zero-shot, Use of domain labels=false2025.05 | 41.3 | — | |
| Point-MoE-LParams=100M, Activated=60M, Training Setting=Multi-dataset Joint Training, Precise Evaluator=false2025.05 | 41.3 | — | |
| Point-MoE-LTraining strategy=Indoor-only joint training, Evaluation setting=Zero-shot2025.05 | 41 | — | |
| PTv3-LParams=97M, Activated=97M, Training Setting=Multi-dataset Joint Training, Mixed Dataset Batch=true, Layernorm=true, Precise Evaluator=false2025.05 | 40.6 | — | |
| OpenScene (reimple)-3DType=Zero-shot2025.03 | 40.5 | 48.8 | |
| Point-MoE-SParams=59M, Activated=52M, Training Setting=Multi-dataset Joint Training, Precise Evaluator=false2025.05 | 40.1 | — | |
| DMA-text onlyType=Zero-shot2025.03 | 39.8 | 49.5 | |
| PTv3-LTraining Setup=Indoor and Outdoor joint training, Evaluation Protocol=Zero-shot, Use of domain labels=false2025.05 | 39.1 | — | |
| PTv3-LTraining strategy=Indoor-only joint training, Evaluation setting=Zero-shot2025.05 | 39 | — | |
| PTv3-LParams=97M, Activated=97M, Training Setting=Multi-dataset Joint Training, Mixed Dataset Batch=true, Precise Evaluator=false2025.05 | 38.4 | — | |
| PTv3-ScanNetParams=46M, Activated=46M, Training Setting=Single-dataset Training, Precise Evaluator=false2025.05 | 37.9 | — | |
| PTv3-SParams=46M, Activated=46M, Training Setting=Multi-dataset Joint Training, Mixed Dataset Batch=true, Layernorm=true, Precise Evaluator=false2025.05 | 35.6 | — | |
| PTv3-SParams=46M, Activated=46M, Training Setting=Multi-dataset Joint Training, Mixed Dataset Batch=true, Precise Evaluator=false2025.05 | 34.9 | — | |
| MSeg VotingType=Zero-shot2025.03 | 33.4 | — | |
| PTv3-LParams=97M, Activated=97M, Training Setting=Multi-dataset Joint Training, Precise Evaluator=false2025.05 | 31 | — | |
| PPT-L*Training strategy=Indoor-only joint training, Evaluation setting=Zero-shot, Note=Uses dataset labels during training2025.05 | 27.8 | — | |
| PTv3-SParams=46M, Activated=46M, Training Setting=Multi-dataset Joint Training, Precise Evaluator=false2025.05 | 27.8 | — | |
| PPT-LParams=98M, Activated=98M, Training Setting=Multi-dataset Joint Training, Precise Evaluator=false2025.05 | 27.8 | — | |
| PPT-L*Training Setup=Indoor and Outdoor joint training, Evaluation Protocol=Zero-shot, Use of domain labels=true2025.05 | 24.9 | — | |
| PPT-SParams=47M, Activated=47M, Training Setting=Multi-dataset Joint Training, Precise Evaluator=false2025.05 | 24.3 | — | |
| PTv3-Structured3DParams=46M, Activated=46M, Training Setting=Single-dataset Training, Precise Evaluator=false2025.05 | 19.5 | — | |
| PTv3-S3DISParams=46M, Activated=46M, Training Setting=Single-dataset Training, Precise Evaluator=false2025.05 | 5.2 | — | |
| DCM-NetType=Fully-sup.2025.03 | — | 66.2 | |
| LGroundType=Fully-sup.2025.03 | — | 67.2 | |
| ScanCompleteType=Fully-sup.2025.03 | — | 44.9 | |
| TangentConvType=Fully-sup.2025.03 | — | 46.8 | |
| TextureNetType=Fully-sup.2025.03 | — | 63 |