Generative Adversarial Networks: Basic Concepts and Applications
摘要
The Generative Adversarial Networks (GAN) is newer entity in deep learning networks with the unique power of data creation. It is probably the most innovative and dynamic machine learning technique. GAN consists of a generator (G) and a discriminator (D) that compete with each other and ultimately, a synthetic image is produced indistinguishable from the original one. There are many variants of GAN with unique features. The large number of synthetic images generated by GANs help to train the neural network. Moreover, GAN can make high-quality realistic images and can be used for different purposes. In this chapter, the basic concept and applications of GAN have been discussed.