Data Scarcity Solution: S-CycleGAN for CT-to-Ultrasound Translation

This paper defines a new, practical setting for Instance-Incremental Learning, focusing on cost-effective model promotion and resistance to catastrophic forgetting