Semantic Segmentation on ISIC
71.1mIoUMPA
Evaluation Results
| Method | Links | |
|---|---|---|
| MPAsource-training=false2026.02 | 71.1 | |
| INSID3Encoder=DINOv3, #Param=304 M, Training Protocol=Training free: Unsupervised pre-training, Shot count=52026.03 | 63.9 | |
| DMTNet (TaP)Version=TaP, Shot count=15-shot2025.12 | 63.86 | |
| DMTNet (TaP)Version=TaP, Shot count=10-shot2025.12 | 61.14 | |
| DMTNet (Decoder FT)Version=Decoder FT, Shot count=15-shot2025.12 | 59.7 | |
| DMTNet (Vanilla)Version=Vanilla, Shot count=15-shot2025.12 | 58.89 | |
| GF-SAM† + our debiasEncoder=DINOv3 + SAM, #Param=945 M, Training Protocol=Training free: Mask-supervised pre-training, Shot count=52026.03 | 58.2 | |
| DMTNet (Decoder FT)Version=Decoder FT, Shot count=10-shot2025.12 | 57.68 | |
| DMTNet (Vanilla)Version=Vanilla, Shot count=10-shot2025.12 | 57.13 | |
| GF-SAM†Encoder=DINOv3 + SAM, #Param=945 M, Training Protocol=Training free: Mask-supervised pre-training, Shot count=52026.03 | 56.7 | |
| DMTNet (TaP)Version=TaP, Shot count=5-shot2025.12 | 56.03 | |
| GF-SAMEncoder=DINOv2 + SAM, #Param=945 M, Training Protocol=Training free: Mask-supervised pre-training, Shot count=52026.03 | 55.2 | |
| TAVPSAM-based=true2026.02 | 54.9 | |
| INSID3Encoder=DINOv3, #Param=304 M, Training Protocol=Training free, Supervision Type=Unsupervised pre-training2026.03 | 54.4 | |
| DMTNet (Decoder FT)Version=Decoder FT, Shot count=5-shot2025.12 | 53.98 | |
| DMTNet (Vanilla)Version=Vanilla, Shot count=5-shot2025.12 | 53.77 | |
| DMTNet (TaP)Version=TaP, Shot count=3-shot2025.12 | 53.27 | |
| GF-SAM† + debiasEncoder=DINOv3 + SAM, #Param=945 M, Training Protocol=Training free, Supervision Type=Mask-supervised pre-training, Debias=True2026.03 | 51.8 | |
| GF-SAM†Encoder=DINOv3 + SAM, #Param=945 M, Training Protocol=Training free, Supervision Type=Mask-supervised pre-training2026.03 | 50.9 | |
| DMTNet (Decoder FT)Version=Decoder FT, Shot count=3-shot2025.12 | 50.7 | |
| DMTNet (Vanilla)Version=Vanilla, Shot count=3-shot2025.12 | 50.01 | |
| GF-SAMEncoder=DINOv2 + SAM, #Param=945 M, Training Protocol=Training free, Supervision Type=Mask-supervised pre-training2026.03 | 48.7 | |
| APSegSAM-based=true2026.02 | 45.4 | |
| SegGPTEncoder=ViT, #Param=354 M, Training Protocol=Task-specific fine-tuning: Semantic + mask supervision, Shot count=52026.03 | 45.2 | |
| MatcherEncoder=DINOv2 + SAM, #Param=945 M, Training Protocol=Training free, Supervision Type=Mask-supervised pre-training2026.03 | 38.6 | |
| SegGPTEncoder=ViT, #Param=354 M, Training Protocol=Task-specific fine-tuning, Supervision Type=Semantic + mask supervision2026.03 | 37.5 | |
| MatcherEncoder=DINOv2 + SAM, #Param=945 M, Training Protocol=Training free: Mask-supervised pre-training, Shot count=52026.03 | 35 | |
| DiffewSEncoder=Stable Diffusion, #Param=890 M, Training Protocol=Task-specific fine-tuning: Semantic + mask supervision, Shot count=52026.03 | 32.7 | |
| SINEEncoder=DINOv2, #Param=373 M, Training Protocol=Task-specific fine-tuning: Semantic + mask supervision, Shot count=52026.03 | 28.6 | |
| DiffewSEncoder=Stable Diffusion, #Param=890 M, Training Protocol=Task-specific fine-tuning, Supervision Type=Semantic + mask supervision2026.03 | 27.8 | |
| SINEEncoder=DINOv2, #Param=373 M, Training Protocol=Task-specific fine-tuning, Supervision Type=Semantic + mask supervision2026.03 | 25.8 | |
| SegICEncoder=DINOv2, #Param=310 M, Training Protocol=Task-specific fine-tuning, Supervision Type=Semantic + mask supervision2026.03 | 25.3 | |
| PerSAMSAM-based=true2026.02 | 23.9 | |
| PerSAMEncoder=SAM, #Param=640 M, Training Protocol=Training free, Supervision Type=Mask-supervised pre-training2026.03 | 23.9 | |
| SegIC (COCO)Encoder=DINOv2, #Param=310 M, Training Protocol=Task-specific fine-tuning, Supervision Type=Semantic + mask supervision, dataset split=COCO2026.03 | 22.5 |