AI Hub
All papers
Architecture

Generative Adversarial Networks

University of Montreal·June 10, 2014

Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza

View on arXiv

TL;DR

Introduces GANs, training a generator against a discriminator to produce realistic synthetic data.

Why it matters

Dominated generative imaging for years and produced the first strikingly realistic synthetic images.

Related terms