AIM++ (256x256, 1.5m parameters, 10% labeled data, no pretraining)

Reported on 2 benchmarks across 1 task · 1 paper · 1 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Medical2 results

Medical Image SegmentationonLesion Segmentation on ISIC 2018
Dice Score· 2024-01-25
0.85
SOTA
Inconsistency Masks: Removing the Uncertainty from Input-Pseudo-Label Pairs arXiv:2401.14387
Medical Image SegmentationonISIC 2018
mean Dice· 2024-01-25
0.85
best: 0.9253 (Polar Res-U-Net++)
Inconsistency Masks: Removing the Uncertainty from Input-Pseudo-Label Pairs arXiv:2401.14387