Abnormality Classification on IU-Xray (test)
0.9079Average AUCFrozen (PET)
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Frozen (PET)Trainable Parameters=2.37M, Efficiency (%)=2.512025.12 | 0.9079 | — | — | — | — | — | — | 0.5645 | 0.3388 | |
| LoRA (PET)Trainable Parameters=2.37M, Efficiency (%)=2.512025.12 | 0.9027 | — | — | — | — | — | — | 0.5635 | 0.3037 | |
| Adapter (PET)Trainable Parameters=2.37M, Efficiency (%)=2.512025.12 | 0.9009 | — | — | — | — | — | — | 0.5634 | 0.2938 | |
| BitFit (PET)Trainable Parameters=2.37M, Efficiency (%)=2.512025.12 | 0.8916 | — | — | — | — | — | — | 0.5538 | 0.3028 | |
| PPKED2021.06 | 0.8 | 0.91 | 0.87 | 0.92 | 0.85 | 0.69 | 0.83 | — | — | |
| DenseNet+KG2021.06 | 0.79 | 0.89 | 0.86 | 0.91 | 0.84 | 0.64 | 0.81 | — | — | |
| TieNet2021.06 | 0.78 | 0.79 | 0.73 | 0.85 | 0.71 | 0.66 | 0.75 | — | — | |
| DenseNet2021.06 | 0.78 | 0.89 | 0.84 | 0.87 | 0.82 | 0.6 | 0.8 | — | — | |
| Full Fine-tuningTrainable Parameters=94.3M, Efficiency (%)=100.02025.12 | 0.7701 | — | — | — | — | — | — | 0.4623 | 0.3273 | |
| Vision-OnlyTrainable Parameters=1.06M, Efficiency (%)=1.122025.12 | 0.6527 | — | — | — | — | — | — | 0.2533 | 0.0488 |