Semantic Segmentation on ScanNet++
62.4Mean IoU (mIoU)TrianguLang
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| TrianguLangSetting=In-domain, Training Data=ScanNet++, Prompt Type=text-only prompts2026.03 | 62.4 | — | 77.4 | — | |
| MV-SAMSetting=In-domain, Training Data=ScanNet++, Prompt Type=12 click prompts2026.03 | 51 | — | 69.4 | — | |
| MV-SAMSetting=Large-scale, Training Data=SA-1B, Prompt Type=12 click prompts2026.03 | 48.9 | — | 63.5 | — | |
| S2GSinference_mode=Zero-shot, input_views=322026.03 | 41.67 | — | — | — | |
| Fit3DBackbone=ViT-Base2026.02 | 34.85 | 84.9 | 44.83 | — | |
| SnDBackbone=ViT-Base2026.02 | 34.07 | 84.77 | 44 | — | |
| SIU3Rinference_mode=Zero-shot, input_views=322026.03 | 33.51 | — | — | — | |
| DINOv2Backbone=ViT-Base2026.02 | 31.95 | 81.85 | 41.69 | — | |
| SnDBackbone=ViT-Small2026.02 | 31.78 | 83.84 | 41.42 | — | |
| Fit3DBackbone=ViT-Small2026.02 | 31.77 | 83.34 | 41.09 | — | |
| MEFBackbone=ViT-Base2026.02 | 31.63 | 82.23 | 41.5 | — | |
| DINOv2Backbone=ViT-Small2026.02 | 29.54 | 80.23 | 39.11 | — | |
| TrianguLangSetting=Cross-domain, Training Data=uCo3D, Prompt Type=text-only prompts2026.03 | 27.9 | — | 68.5 | — | |
| MEFBackbone=ViT-Small2026.02 | 27.44 | 79.45 | 36.77 | — | |
| MV-SAMSetting=Cross-domain, Training Data=uCo3D, Prompt Type=12 click prompts2026.03 | 19.4 | — | 25.1 | — | |
| Feature-3DGSBase Model=true2026.06 | — | — | — | 22.47 | |
| IGGTBase Model=true2026.06 | — | — | — | 31.31 | |
| LSegBase Model=true2026.06 | — | — | — | 22.61 | |
| LSMmode=Multi-View2026.06 | — | — | — | 17.88 | |
| OpenSegBase Model=true2026.06 | — | — | — | 13.92 | |
| Pano3Dbackbone=MUSt3R2026.06 | — | — | — | 46.61 | |
| Pano3Dbackbone=Pi32026.06 | — | — | — | 52.54 |