Medical Images Augmentation via GAN Image patches segmentation using Yolo with Neural Style

Abstract

The problem that paper is solving relates to Low Model performance due To Fewer Images to train. Medical field is the area in which we encounter less amount of training data due to rare diseases like Lung Cancer, Histopathological Cancer, Covid 19 etc.

The problem with small datasets is that models trained with them suffer from the problem of over fitting.

Image Augmentation is another way we can reduce over fitting on models, where we increase the number of training images using information only in our training image.

 Key area of the Research paper is to get Image Augmentation with GAN using YOLO and Neural Style.