Consistency learning with dynamic weighting and class-agnostic regularization for semi-supervised medical image segmentation

Jiawei Su, Zhiming Luo, Sheng Lian, Dazhen Lin, Shaozi Li

Abstract: Highlights•We propose a dynamic weight sampling (DWS) to encourage high-confidence learning.•Relationship between pixels is used as a robust signal to capture valuable features.•Our method achieves state of the art for semi-supervised medical image segmentation.