SegWithU is a post-hoc framework for uncertainty estimation in medical image segmentation. It augments a frozen pretrained segmentation backbone with a lightweight uncertainty head, modeling uncertainty as perturbation energy with rank-1 posterior probes. It produces voxel-wise uncertainty maps for calibration and error detection in a single forward pass.

This page is only used to distribute the checkpoints. For code and docs please visit our GitHub repository or the project page.