Abstract: Data augmentation is an effective way to improve the generalization of deep learning models. However, the underlying augmentation methods mainly rely on handcrafted operations, such as ...

Abstract: Deep learning has attracted wide attention recently because of its excellent feature representation ability and end-to-end automatic learning method. Especially in clinical medical imaging ...