3D Semantic Segmentation on S3DIS Area5
76mIoUSonata
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| SonataParam. (M)=124.8, Evaluation Protocol=Full Fine-tuning, Decoder Setting=With decoder2026.04 | 76 | 81.6 | 93 | |
| Sonata + PointGSTParam. (M)=1.05 (0.97%), Evaluation Protocol=PEFT methods for point cloud, Decoder Setting=Without decoder2026.04 | 75 | 81.4 | 92.9 | |
| Sonata + PointTPAParam. (M)=1.18 (1.09%), Evaluation Protocol=PEFT methods for point cloud, Decoder Setting=Without decoder2026.04 | 74.9 | 81.7 | 92.9 | |
| Sonata + DAPTParam. (M)=1.14 (1.06%), Evaluation Protocol=PEFT methods for point cloud, Decoder Setting=Without decoder2026.04 | 74.6 | 81.2 | 92.7 | |
| SonataParam. (M)=108.5, Evaluation Protocol=Full Fine-tuning, Decoder Setting=Without decoder2026.04 | 74.5 | 80.3 | 93.1 | |
| PTv3-PPT (sup.)Param. (M)=124.8, Evaluation Protocol=Full Fine-tuning2026.04 | 74.3 | 80.1 | 92 | |
| Sonata + LoRAParam. (M)=0.89 (0.82%), Evaluation Protocol=General PEFT methods, Decoder Setting=Without decoder2026.04 | 74 | 81.1 | 92.5 | |
| Sonata + AdapterParam. (M)=1.90 (1.76%), Evaluation Protocol=General PEFT methods, Decoder Setting=Without decoder2026.04 | 73.8 | 81.7 | 91.6 | |
| Sonata + BitFitParam. (M)=0.12 (0.11%), Evaluation Protocol=General PEFT methods, Decoder Setting=Without decoder2026.04 | 73.8 | 81.2 | 92.2 | |
| PTv3Param. (M)=124.8, Evaluation Protocol=Training from scratch2026.04 | 73.4 | 78.9 | 91.7 | |
| SonataParam. (M)=0.02 (0.02%), Evaluation Protocol=linear probing2026.04 | 73 | 80.9 | 90.8 | |
| Sonata + Prefix TuningParam. (M)=0.08 (0.08%), Evaluation Protocol=General PEFT methods, Decoder Setting=Without decoder2026.04 | 73 | 82.1 | 91 | |
| Sonata + RandLoRAParam. (M)=0.71 (0.66%), Evaluation Protocol=General PEFT methods, Decoder Setting=Without decoder2026.04 | 73 | 81.7 | 91.2 | |
| Sonata + VeRAParam. (M)=0.08 (0.08%), Evaluation Protocol=General PEFT methods, Decoder Setting=Without decoder2026.04 | 72.9 | 80.8 | 91 | |
| Sonata + IDPTParam. (M)=2.61 (2.42%), Evaluation Protocol=PEFT methods for point cloud, Decoder Setting=Without decoder2026.04 | 72 | 81.2 | 90.5 | |
| PointNeXt-XLParam. (M)=41.6, Evaluation Protocol=Training from scratch2026.04 | 70.5 | — | 90.6 | |
| Point-MoE-LParams=100M, Activated=59M, Training Setting=Multi-dataset Joint Training2025.05 | 69.5 | — | — | |
| PTv3-SParams=46M, Activated=46M, Training Setting=Multi-dataset Joint Training2025.05 | 69.4 | — | — | |
| Point-MoE-SParams=59M, Activated=52M, Training Setting=Multi-dataset Joint Training, Precise Evaluator=false2025.05 | 68.6 | — | — | |
| Point-MoE-LParams=100M, Activated=60M, Training Setting=Multi-dataset Joint Training, Precise Evaluator=false2025.05 | 68.5 | — | — | |
| Point-MoE-SParams=59M, Activated=52M, Training Setting=Multi-dataset Joint Training2025.05 | 68 | — | — | |
| PTv3-LParams=97M, Activated=97M, Training Setting=Multi-dataset Joint Training, Mixed Dataset Batch=true, Layernorm=true, Precise Evaluator=false2025.05 | 67.8 | — | — | |
| PTv3-S3DISParams=46M, Activated=46M, Training Setting=Single-dataset Training2025.05 | 67.6 | — | — | |
| PTv3-LParams=97M, Activated=97M, Training Setting=Multi-dataset Joint Training2025.05 | 67.4 | — | — | |
| PTv3-LParams=97M, Activated=97M, Training Setting=Multi-dataset Joint Training, Mixed Dataset Batch=true, Precise Evaluator=false2025.05 | 67.2 | — | — | |
| PTv3-SParams=46M, Activated=46M, Training Setting=Multi-dataset Joint Training, Mixed Dataset Batch=true, Layernorm=true, Precise Evaluator=false2025.05 | 66.9 | — | — | |
| SparseUNetParam. (M)=39.2, Evaluation Protocol=Training from scratch2026.04 | 66.3 | 72.5 | 89.8 | |
| PTv3-S3DISParams=46M, Activated=46M, Training Setting=Single-dataset Training, Precise Evaluator=false2025.05 | 66.1 | — | — | |
| PPT-L*Params=98M, Activated=98M, Training Setting=Multi-dataset Joint Training2025.05 | 65 | — | — | |
| PPT-S*Params=47M, Activated=47M, Training Setting=Multi-dataset Joint Training2025.05 | 64.7 | — | — | |
| PTv3-SParams=46M, Activated=46M, Training Setting=Multi-dataset Joint Training, Mixed Dataset Batch=true, Precise Evaluator=false2025.05 | 62.5 | — | — | |
| PPT-LParams=98M, Activated=98M, Training Setting=Multi-dataset Joint Training, Precise Evaluator=false2025.05 | 62 | — | — | |
| PPT-SParams=47M, Activated=47M, Training Setting=Multi-dataset Joint Training, Precise Evaluator=false2025.05 | 61 | — | — | |
| OA-CNNsParams=52M, Activated=52M, Training Setting=Multi-dataset Joint Training2025.05 | 60.6 | — | — | |
| PTv3-SParams=46M, Activated=46M, Training Setting=Multi-dataset Joint Training, Precise Evaluator=false2025.05 | 44.1 | — | — | |
| PTv3-LParams=97M, Activated=97M, Training Setting=Multi-dataset Joint Training, Precise Evaluator=false2025.05 | 40.3 | — | — | |
| PTv3-Matterport3DParams=46M, Activated=46M, Training Setting=Single-dataset Training, Precise Evaluator=false2025.05 | 9.5 | — | — | |
| PTv3-ScanNetParams=46M, Activated=46M, Training Setting=Single-dataset Training2025.05 | 8.8 | — | — | |
| PTv3-ScanNetParams=46M, Activated=46M, Training Setting=Single-dataset Training, Precise Evaluator=false2025.05 | 8.5 | — | — | |
| PTv3-Structured3DParams=46M, Activated=46M, Training Setting=Single-dataset Training2025.05 | 7.1 | — | — | |
| PTv3-Structured3DParams=46M, Activated=46M, Training Setting=Single-dataset Training, Precise Evaluator=false2025.05 | 7 | — | — |