Semantic Segmentation on ADE
46.4mIoUMultiMAE
Evaluation Results
| Method | Links | |
|---|---|---|
| MultiMAEBackbone=ViT-Base/16, Pre-training Data=IN1K2022.10 | 46.4 | |
| MAEBackbone=ViT-Base/16, Pre-training Data=IN1K2022.10 | 46.1 | |
| DINOBackbone=ViT-Base/16, Pre-training Data=IN1K2022.10 | 44.7 | |
| CroCoBackbone=ViT-Base/16, Pre-training Data=Habitat2022.10 | 40.6 | |
| MAEBackbone=ViT-Base/16, Pre-training Data=Habitat2022.10 | 40.3 | |
| CoCoOp+CAKIBase Model=SEEM-Large2026.05 | 26.4 | |
| CoCoOpBase Model=SEEM-Large2026.05 | 25.6 | |
| CoCoOp+CAKIBase Model=SEEM-Tiny2026.05 | 21.6 | |
| ProxyCLIP + RAG-OVSVenue=This Paper2025.12 | 21.1 | |
| ProxyCLIPVenue=ECCV'242025.12 | 20.2 | |
| CoCoOpBase Model=SEEM-Tiny2026.05 | 20.2 | |
| CoOp+CAKIBase Model=SEEM-Tiny2026.05 | 19.5 | |
| CoOp+CAKIBase Model=SEEM-Large2026.05 | 19.3 | |
| NACLIP + RAG-OVSVenue=This Paper2025.12 | 19.1 | |
| CLIPtrace + RAG-OVSVenue=This Paper2025.12 | 18.5 | |
| SCLIP + RAG-OVSVenue=This Paper2025.12 | 18.1 | |
| CoOpBase Model=SEEM-Large2026.05 | 18.1 | |
| DiffTPT + HisTPTModel Scale=Tiny2024.10 | 17.9 | |
| TPT + HisTPTModel Scale=Tiny2024.10 | 17.6 | |
| CoOpBase Model=SEEM-Tiny2026.05 | 17.6 | |
| DiffTPT + HisTPTModel Scale=Large2024.10 | 17.4 | |
| CLIP-Surgery + RAG-OVSVenue=This Paper2025.12 | 17.4 | |
| NACLIPVenue=WACV'252025.12 | 17.4 | |
| HisTPTModel Scale=Tiny2024.10 | 17.3 | |
| TPT + HisTPTModel Scale=Large2024.10 | 17.1 | |
| CLIPtraceVenue=ECCV'242025.12 | 17 | |
| HisTPTModel Scale=Large2024.10 | 16.9 | |
| DiffTPTModel Scale=Tiny2024.10 | 16.3 | |
| TPTModel Scale=Tiny2024.10 | 16.1 | |
| DiffTPTModel Scale=Large2024.10 | 16.1 | |
| SCLIPVenue=ECCV'242025.12 | 16.1 | |
| CLIP-SurgeryVenue=Pattern Recognition2025.12 | 15.8 | |
| TPTModel Scale=Large2024.10 | 15.7 | |
| GroupViTVenue=CVPR'222025.12 | 15.3 | |
| SEEMModel Scale=Large2024.10 | 15.2 | |
| SEEM-Large2026.05 | 15.2 | |
| RecoVenue=NeurIPS'222025.12 | 14.8 | |
| SEEMModel Scale=Tiny2024.10 | 14.6 | |
| SEEM-Tiny2026.05 | 14.6 | |
| OVDiffVenue=ECCV'242025.12 | 14.1 | |
| MaskCLIPVenue=ECCV'222025.12 | 11.9 | |
| CLIPVenue=ICML'212025.12 | 2.9 |