Image Classification on Camelyon 17 (test)
92.3AccuracyTBF
Evaluation results
| Method | Links | |
|---|---|---|
| TBFEvaluation Protocol=Fine-tuned, Random Perturbation=true2025.12 | 92.3 | |
| TBFEvaluation Protocol=Fine-tuned, Tool Selection=All modality-specific tools2025.12 | 92.1 | |
| TBFEvaluation Protocol=Fine-tuned, Random Perturbation=false, alpha=12025.12 | 90.2 | |
| EfficientNetEvaluation Protocol=Fine-tuned2025.12 | 88.6 | |
| Y-NetEvaluation Protocol=Fine-tuned2025.12 | 88.2 | |
| asymmetric knowledge distillation frameworkDistortion type=90% random masking (RM), Evaluation protocol=Transfer learning2026.04 | 76.03 | |
| asymmetric knowledge distillation frameworkDistortion type=Gaussian blur with kernel size 37 and σ = 9 (GB), Evaluation protocol=Transfer learning2026.04 | 74.77 | |
| CIDistortion type=Gaussian blur with kernel size 37 and σ = 9 (GB), Evaluation protocol=Transfer learning2026.04 | 73.88 | |
| LlavaMed FTEvaluation Protocol=Fine-tuned2025.12 | 66.2 | |
| asymmetric knowledge distillation frameworkDistortion type=Gaussian noise with σ = 0.5 (GN), Evaluation protocol=Transfer learning2026.04 | 65.24 | |
| CIDistortion type=Gaussian noise with σ = 0.5 (GN), Evaluation protocol=Transfer learning2026.04 | 62.27 | |
| BaselineDistortion type=Gaussian blur with kernel size 37 and σ = 9 (GB), Evaluation protocol=Transfer learning2026.04 | 59.73 | |
| BaselineDistortion type=90% random masking (RM), Evaluation protocol=Transfer learning2026.04 | 56.2 | |
| BaselineDistortion type=Gaussian noise with σ = 0.5 (GN), Evaluation protocol=Transfer learning2026.04 | 55.86 | |
| CIDistortion type=90% random masking (RM), Evaluation protocol=Transfer learning2026.04 | 55.84 | |
| VisProgEvaluation Protocol=Zero-shot2025.12 | 50.4 | |
| MedGemmaEvaluation Protocol=Zero-shot2025.12 | 50 | |
| MedGemmaEvaluation Protocol=Zero-shot, Tool Prompts=Included2025.12 | 50 | |
| LlavaMedEvaluation Protocol=Zero-shot2025.12 | 50 |