Semantic Segmentation on ADE20K (Scale-Specific mIoU Reporting)
49.8mIoU (Overall)UniMatch V2 (published)
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| UniMatch V2 (published)Encoder=DINOv2-B2026.07 | 49.8 | — | — | — | — | — | |
| PixCon (ours)Encoder=DINOv2-B2026.07 | 49.23 | — | — | — | — | — | |
| UniMatch V2 (our repro)Encoder=DINOv2-B2026.07 | 49.1 | — | — | — | — | — | |
| UniMatch V2 (published)Encoder=DINOv2-S2026.07 | 44.4 | — | — | — | — | — | |
| SemiVLEncoder=CLIP-B2026.07 | 39.4 | — | — | — | — | — | |
| UniMatchEncoder=CLIP-B2026.07 | 38 | — | — | — | — | — | |
| UniMatchEncoder=RN-1012026.07 | 34.6 | — | — | — | — | — | |
| CLIP-Guided SAM (labelled-only)Unlabelled data=No, Image-level prompts=Yes, Encoder=CLIP-B+SAM-B, Number of Parameters=197M2026.05 | — | 47.9 | 51.6 | 55.2 | 57.7 | 61.1 | |
| SemiVLUnlabelled data=Yes, Image-level prompts=No, Encoder=CLIP-Base, Number of Parameters=88M2026.05 | — | 33.7 | 35.1 | 37.2 | 39.4 | — | |
| UniMatchV2Unlabelled data=Yes, Image-level prompts=No, Encoder=DINOv2-B, Number of Parameters=97.5M2026.05 | — | 38.7 | 45 | 46.7 | 49.8 | 52 | |
| UniMatchV2 (labelled-only)Unlabelled data=No, Image-level prompts=No, Encoder=DINOv2-B, Number of Parameters=97.5M2026.05 | — | 26.1 | 39.3 | 42.8 | 46.4 | 49 |