Semantic Segmentation on ClinicDB (test)
93.63DSC (%)CRIS
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| CRISTraining Strategy=E2E fine-tuning, Trainable Parameters=147M2024.05 | 93.63 | 88.74 | 12.36 | |
| CRISEvaluation Protocol=End-to-End fine-tuning, Parameters=147M2025.12 | 93.63 | 88.74 | — | |
| Telescopic AdaptersEvaluation Protocol=Adapter Fine-Tuning, Parameters=613k2025.12 | 91.67 | 85.19 | — | |
| CLIPSeg DA VLCEvaluation Protocol=Adapter Fine-Tuning, Parameters=3.2M2025.12 | 89.63 | 82.89 | — | |
| CLIPSeg SATraining Strategy=Adapter fine-tuning, Trainable Parameters=4.2M2024.05 | 89.04 | 81.93 | 18.03 | |
| CLIPSeg SA VLCEvaluation Protocol=Adapter Fine-Tuning, Parameters=4.2M2025.12 | 89.04 | 81.95 | — | |
| CLIPSeg DATraining Strategy=Adapter fine-tuning, Trainable Parameters=3M2024.05 | 88.73 | 81.84 | 18.76 | |
| CLIPSegTraining Strategy=E2E fine-tuning, Trainable Parameters=150M2024.05 | 88.58 | 81.51 | 19.3 | |
| CLIPSegEvaluation Protocol=End-to-End fine-tuning, Parameters=150M2025.12 | 88.58 | 81.51 | — | |
| I-MedSAMEvaluation Protocol=LoRA Fine-Tuning, Parameters=1.6M2025.12 | 88.12 | 79.91 | — | |
| SANTraining Strategy=Adapter fine-tuning, Trainable Parameters=8.4M2024.05 | 81.36 | 72.61 | 38.42 | |
| SANEvaluation Protocol=Adapter Fine-Tuning, Parameters=8.4M2025.12 | 81.36 | 72.61 | — | |
| CLIPSegEvaluation Protocol=Zero-Shot, Parameters=150M2025.12 | 36.58 | 27.54 | — |