Generative adversarial networks

GANs are evolving rather quickly, and are receiving a considerable amount of attention from the research community. Yann LeCun's comment, expressing that the GAN framework is the most interesting idea in the last 10 years of machine learning, shows evidence of the perceived importance of the framework.

The following is a figure representing applications of the GAN framework:

Source: Generative Adversarial Nets (https://arxiv.org/abs/1406.2661)

The GAN framework has been widely used to generate data from many domains. Examples of data generation with GANs include text-to-image synthesis, image super-resolution, ...

Get Hands-On Generative Adversarial Networks with Keras now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.