Mitochondria Segmentation on Human → Lucchi++
95.1Dice ScoreFully-Supervised Model
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Fully-Supervised ModelTypes=Oracle2026.02 | 95.1 | 90.1 | 85.8 | |
| Prefer-DAS (LPO)+Segmentation Mode=Interactive, Adaptation Type=WDA, Training Point Prompts=15%2026.02 | 94.7 | 89.2 | 84.4 | |
| Prefer-DAS (LPO)Segmentation Mode=Automatic, Adaptation Type=WDA, Training Point Prompts=15%2026.02 | 94.1 | 88.3 | 82.3 | |
| Prefer-DAS (SLPO)+Segmentation Mode=Interactive, Adaptation Type=WDA, Training Point Prompts=15%2026.02 | 93.9 | 87.9 | 83 | |
| Prefer-DAS (UPO)+Segmentation Mode=Interactive, Adaptation Type=WDA, Training Point Prompts=15%2026.02 | 93.4 | 87.1 | 82.5 | |
| Prefer-DAS (SLPO)Segmentation Mode=Automatic, Adaptation Type=WDA, Training Point Prompts=15%2026.02 | 93.2 | 86.8 | 81.3 | |
| Prefer-DAS (UPO)Segmentation Mode=Automatic, Adaptation Type=WDA, Training Point Prompts=15%2026.02 | 92.6 | 85.6 | 80.9 | |
| Prefer-DAS (UPO)Segmentation Mode=Automatic, Adaptation Type=UDA, Training Point Prompts=0%2026.02 | 92.1 | 84.7 | 79.6 | |
| Prompt-DAS+Segmentation Mode=Interactive, Adaptation Type=WDA, Training Point Prompts=15%2026.02 | 88.7 | 79.1 | 74.8 | |
| Prompt-DASSegmentation Mode=Automatic, Adaptation Type=WDA, Training Point Prompts=15%2026.02 | 88.2 | 78.3 | 74.8 | |
| Prompt-DAS (UDA)Segmentation Mode=Automatic, Adaptation Type=UDA, Training Point Prompts=0%2026.02 | 87.3 | 76.9 | 73.1 | |
| Prefer-DAS (Source)+Segmentation Mode=Interactive, Adaptation Type=NoAdapt, Training Point Prompts=0%2026.02 | 87 | 75.5 | 69.4 | |
| Med-SAM Adapter+Segmentation Mode=Interactive, Adaptation Type=NoAdapt, Training Point Prompts=0%, Fine-tuned on source=true2026.02 | 86.8 | 74 | 56.8 | |
| CAFASegmentation Mode=Automatic, Adaptation Type=UDA, Training Point Prompts=0%2026.02 | 86.7 | 76.7 | 70.1 | |
| DAMT-NetSegmentation Mode=Automatic, Adaptation Type=UDA, Training Point Prompts=0%2026.02 | 86.6 | 75.5 | 73 | |
| WDA-NetSegmentation Mode=Automatic, Adaptation Type=WDA, Training Point Prompts=15%2026.02 | 86.3 | 75.6 | 70.8 | |
| Prefer-DAS (Source)Segmentation Mode=Automatic, Adaptation Type=NoAdapt, Training Point Prompts=0%2026.02 | 85.6 | 73.5 | 66.8 | |
| WDA-Net (UDA)Segmentation Mode=Automatic, Adaptation Type=UDA, Training Point Prompts=0%2026.02 | 85.6 | 74.3 | 71.9 | |
| DA-ISCSegmentation Mode=Automatic, Adaptation Type=UDA, Training Point Prompts=0%2026.02 | 85.2 | 72.8 | 64.1 | |
| WeSAM+Segmentation Mode=Interactive, Adaptation Type=WDA, Training Point Prompts=15%, Fine-tuned on source and target=true2026.02 | 83.2 | 69 | 63 | |
| UALRSegmentation Mode=Automatic, Adaptation Type=UDA, Training Point Prompts=0%2026.02 | 81.3 | 68.1 | 63.2 | |
| SAM+Segmentation Mode=Interactive, Adaptation Type=NoAdapt, Training Point Prompts=0%2026.02 | 81.1 | 68.1 | 67.6 | |
| SAM-Med2D+Segmentation Mode=Interactive, Adaptation Type=NoAdapt, Training Point Prompts=0%2026.02 | 78.8 | 63.4 | 35.9 | |
| Med-SAM AdapterSegmentation Mode=Automatic, Adaptation Type=NoAdapt, Training Point Prompts=0%, Fine-tuned on source=true2026.02 | 64.6 | 44.8 | 19.5 | |
| SAM-Med2DSegmentation Mode=Automatic, Adaptation Type=NoAdapt, Training Point Prompts=0%2026.02 | 23.1 | 12.7 | 2.6 | |
| SAMSegmentation Mode=Automatic, Adaptation Type=NoAdapt, Training Point Prompts=0%2026.02 | 21.5 | 10.3 | 11.3 |