Out-of-Distribution Detection on BloodMNIST
90.21ID AccuracyDeep Ensembles
Evaluation Results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Deep EnsemblesM=52026.05 | 90.21 | 43.95 | 59.66 | 51.34 | 51.65 | 98.48 | 92.5 | 90.73 | 93.9 | |
| VI-EDL2026.05 | 89.71 | 72.01 | 76.25 | 77.48 | 75.25 | 33.43 | 25.34 | 53.43 | 37.4 | |
| I-EDLVenue=ICML 20232026.05 | 89.65 | 61.32 | 53.23 | 37.77 | 50.77 | 88.62 | 98.74 | 92.05 | 93.14 | |
| NatPNVenue=ICLR 20222026.05 | 89.21 | 62.82 | 63.25 | 68.44 | 64.84 | 98.87 | 96.15 | 73.78 | 89.6 | |
| F-EDLVenue=NeurIPS 20252026.05 | 86.64 | 41.01 | 53.3 | 52.6 | 48.97 | 99.54 | 93.81 | 93.76 | 95.7 | |
| Re-EDLVenue=TPAMI 20252026.05 | 86 | 57.33 | 59.94 | 52.53 | 56.6 | 94 | 90.98 | 72.52 | 85.83 | |
| EDLVenue=NeurIPS 20182026.05 | 82.72 | 31.3 | 43.01 | 15.33 | 29.88 | 99.58 | 92.22 | 99.51 | 97.1 | |
| SoftmaxLoss=CE2026.05 | 80.18 | 45.05 | 55.53 | 30.39 | 43.66 | 98.05 | 92.69 | 98.02 | 96.25 |