Generative Adversarial Networks (GANs)

GANs are very popular frameworks for generating high-squality data and are immensely used in academia and industry in many domains.

GANs are a powerful class of neural networks used for unsupervised learning. It was developed and introduced by Ian J. Goodfellow in 2014. There are two competing neural network models in a GAN, and they compete with each other to capture and replicate the differences within a dataset.

A discriminator and a generator are both present in GANs. The Generator generates fake data samples (be it an image, audio, etc.) and tries to fool the Discriminator. Discriminator tries to tell the difference between authentic and bogus samples. Because they are both neural networks, the Generator and Discriminator are constantly competing during the training process.