Out-of-Distribution Detection on LIDC-Mal
96AUROCAQA
Evaluation results
| Method | Links | |
|---|---|---|
| AQAParameter=.92026.03 | 96 | |
| PLMParameter=202026.03 | 96 | |
| AQAParameter=.752026.03 | 96 | |
| AQAParameter=.62026.03 | 96 | |
| PLMParameter=502026.03 | 96 | |
| AVGStrategy=Averaging2026.03 | 96 | |
| PLM 20Uncertainty Method=MC Dropout, Patch Size=202026.03 | 96 | |
| AQA .75Uncertainty Method=MC Dropout, Threshold=.752026.03 | 96 | |
| AQA .9Uncertainty Method=MC Dropout, Threshold=.92026.03 | 96 | |
| AQA .9Augmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 95 | |
| PLM 20Augmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 95 | |
| AQA .75Augmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 95 | |
| AQA .6Augmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 95 | |
| PLM 50Augmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 95 | |
| AVGAugmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 95 | |
| PLM 50Uncertainty Method=MC Dropout, Patch Size=502026.03 | 95 | |
| AQA 0.60Uncertainty Method=MC Dropout, Threshold=0.602026.03 | 95 | |
| AVGUncertainty Method=MC Dropout2026.03 | 95 | |
| PLMParameter=102026.03 | 93 | |
| PLM 10Uncertainty Method=MC Dropout, Patch Size=102026.03 | 91 | |
| GMM-IntAugmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 90 | |
| PLM 10Augmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 90 | |
| GMM-AllAugmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 89 | |
| GMM-IntStrategy=GMM-Int2026.03 | 89 | |
| GMM-AllStrategy=GMM-All2026.03 | 86 | |
| GMM-AllUncertainty Method=MC Dropout2026.03 | 86 | |
| GMM-IntUncertainty Method=MC Dropout2026.03 | 86 | |
| GMM-SpaStrategy=GMM-Spa2026.03 | 83 | |
| GMM-SpaAugmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 78 | |
| QFRStrategy=QFR2026.03 | 77 | |
| BCAStrategy=BCA2026.03 | 71 | |
| BCAAugmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 70 | |
| ICAAugmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 70 | |
| ICAStrategy=ICA2026.03 | 70 | |
| QFRAugmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 68 | |
| GMM-SpaUncertainty Method=MC Dropout2026.03 | 67 | |
| ATA .3Augmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 64 | |
| ATAParameter=.32026.03 | 60 | |
| BCAUncertainty Method=MC Dropout2026.03 | 57 | |
| ICAUncertainty Method=MC Dropout2026.03 | 57 | |
| ATA .5Augmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 55 | |
| QFRUncertainty Method=MC Dropout2026.03 | 54 | |
| ATAParameter=.52026.03 | 51 | |
| ATA .7Uncertainty Method=MC Dropout, Threshold=.72026.03 | 51 | |
| ATA .5Uncertainty Method=MC Dropout, Threshold=.52026.03 | 47 | |
| ATA .7Augmentation strategy=Test-Time Augmentation (TTA), Number of bootstrap samples=5002026.03 | 46 | |
| ATAParameter=.72026.03 | 42 | |
| ATA .3Uncertainty Method=MC Dropout, Threshold=.32026.03 | 42 |