Semantic Segmentation on BKAI (test)
92.62DSCCRIS
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| CRISTraining Strategy=E2E fine-tuning, Trainable Parameters=147M2024.05 | 92.62 | 88.3 | 49.8 | |
| CRISEvaluation Protocol=End-to-End fine-tuning, Parameters=147M2025.12 | 92.62 | 88.3 | — | |
| Telescopic AdaptersEvaluation Protocol=Adapter Fine-Tuning, Parameters=613k2025.12 | 88.38 | 81.63 | — | |
| I-MedSAMEvaluation Protocol=LoRA Fine-Tuning, Parameters=1.6M2025.12 | 88.05 | 80.87 | — | |
| CLIPSeg DATraining Strategy=Adapter fine-tuning, Trainable Parameters=3M2024.05 | 87.23 | 79.81 | 70.02 | |
| CLIPSeg DA VLCEvaluation Protocol=Adapter Fine-Tuning, Parameters=3.2M2025.12 | 87.17 | 79.74 | — | |
| CLIPSegTraining Strategy=E2E fine-tuning, Trainable Parameters=150M2024.05 | 85.59 | 77.52 | 87.91 | |
| CLIPSegEvaluation Protocol=End-to-End fine-tuning, Parameters=150M2025.12 | 85.59 | 77.52 | — | |
| CLIPSeg SA VLCEvaluation Protocol=Adapter Fine-Tuning, Parameters=4.2M2025.12 | 83.95 | 75.61 | — | |
| CLIPSeg SATraining Strategy=Adapter fine-tuning, Trainable Parameters=4.2M2024.05 | 83.67 | 75.02 | 87.79 | |
| SANTraining Strategy=Adapter fine-tuning, Trainable Parameters=8.4M2024.05 | 66.26 | 54.58 | 224.37 | |
| SANEvaluation Protocol=Adapter Fine-Tuning, Parameters=8.4M2025.12 | 66.26 | 54.58 | — | |
| CLIPSegEvaluation Protocol=Zero-Shot, Parameters=150M2025.12 | 30.08 | 23.96 | — |