Browse Teach Curate

Generative Adversarial Networks in Deep Learning

Training a generator against a discriminator so that the generator learns to produce samples the discriminator cannot distinguish from real data. Covers the minimax objective, the instability and mode collapse that follow from it, and conditional variants that translate an input image into a different domain or condition.

jeremy Video

Understanding GANs (Generative Adversarial Networks)

Generative Adversarial Networks (GANs) operate within the domain of unsupervised machine learning via a theoretical framework that models generative problems as a two-player minimax game between a ge…