An experiment of the effects of rotatory attention module on many types of Deep Learning networks for image segmentation

Objectives

1. Study on the best position to integrate rotatory attention
2. Study on different ways of calculating loss for models with rotatory attention
3. Study how rotatory attention work with Grad-CAM