Generative Adversarial Networks (GAN): A Survey

Abstract

Recently, Generative Adversarial Networks (GANs) has emerged as a re- search interest in the field of deep learning. GAN works on the analysis of distribution  of ground truth data samples and creating new samples by using that analysis. It workson two models, Generator and Discriminator, trained using backpropagation through a competitive procedure. GAN has gained enormous popularity in the fields of image processing, Vision computation, image synthesis, natural language processing, signal processing. This paper highlights basic architecture of GAN, and different variants of GAN being used byresearchers