Medical Image Synthesis: S-CycleGAN for RUSS and Segmentation

This article presents a novel Decision Boundary-Aware Distillation methodology for Instance-Incremental Learning that requires no access to old data.