OOD-SEG: Exploiting out-of-distribution detection techniques for learning image segmentation from sparse multi-class positive-only annotations

ArXi:2411.09553v4 Announce Type: replace Despite significant advancements, segmentation based on deep neural networks in medical and surgical imaging faces several challenges, two of which we aim to address in this work. First, acquiring complete pixel-level segmentation labels for medical images is time-consuming and requires domain expertise. Second, typical segmentation pipelines cannot detect out-of-distribution (OOD) pixels, leaving them prone to spurious outputs during deployment.