Identity-Aware U-Net: Fine-grained Cell Segmentation via Identity-Aware Representation Learning

ArXi:2604.09702v1 Announce Type: cross Precise segmentation of objects with highly similar shapes remains a challenging problem in dense prediction, especially in scenarios with ambiguous boundaries, overlapping instances, and weak inter-instance visual differences. While conventional segmentation models are effective at localizing object regions, they often lack the discriminative capacity required to reliably distinguish a target object from morphologically similar distractors.