Medical Image Segmentation on ACDC (Dice, Brier, AUROC, AURC)
90.61DiceDeep Ensembles
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Deep EnsemblesType=Multi-pass2026.04 | 90.61 | 0.0107 | 98.13 | 2.5326 | |
| Temperature ScalingType=Single-forward-pass2026.04 | 90.35 | 0.0115 | 97.14 | 4.6798 | |
| DDU-SegType=Single-forward-pass2026.04 | 90.35 | 0.0122 | 97.16 | 4.6628 | |
| SegWithUType=Single-forward-pass2026.04 | 90.35 | 0.0113 | 98.38 | 2.4885 | |
| DUEType=Single-forward-pass2026.04 | 90.24 | 0.0115 | 98.09 | 3.0233 | |
| Monte Carlo DropoutType=Multi-pass2026.04 | 90.1 | 0.0123 | 95.49 | 6.3111 | |
| DUQType=Single-forward-pass2026.04 | 88.71 | 0.0141 | 97.35 | 4.7843 | |
| Test-time AugmentationType=Multi-pass2026.04 | 66.07 | 0.0251 | 98.26 | 8.15 |