Generative adversarial networks - an introduction

Generative adversarial networks (GANs) present a way to learn deep representations without extensively annotated training data. These networks achieve learning through deriving back propagation signals through a competitive process involving a pair of networks. The representations that can be learned by GANs may be used in several applications. GANs have made significant advancements and tremendous performance in numerous applications. The essential applications include semantic image editing, style transfer, image synthesis, image super-resolution and classification. This chapter aims to present an overview of GANs, and its different variants. The chapter attempts to identify GANs' advantages, disadvantages and significant challenges to the successful implementation of GAN in different application areas. Finally, the chapter ends with the conclusion and future aspects.