Introduction

The power of creativity was always the exclusive domain of the human mind. This was one of the facts touted as one of the major differences between the human mind and the artificial intelligence domain. However, in the recent past, deep learning has been making baby steps in the path to being creative. Imagine you were at the Sistine Chapel in the Vatican and were looking up with bewilderment at the frescos immortalized by Michelangelo, wishing your deep learning models were able to recreate something like that. Well, maybe 10 years back, people would have scoffed at your thought. Not anymore, though – deep learning models have made great strides in regenerating immortal works. Applications like these are made possible by a class of networks called Generative Adversarial Networks (GANs).

Many applications have been made possible with GANs. Take a look at the following image:

Figure 7.1: Image translation using GANs

Note:

The preceding...