Skip to footer
Home Research Artificial Intelligence / Machine Learning Metropolis-Hastings Generative Adversarial Networks

Metropolis-Hastings Generative Adversarial Networks


We introduce the Metropolis-Hastings generative adversarial network (MH-GAN), which combines aspects of Markov chain Monte Carlo and GANs. The MH-GAN draws samples from the distribution implicitly defined by a GAN’s discriminator-generator pair, as opposed to sampling in a standard GAN which draws samples from the distribution defined by the generator. It uses the discriminator from GAN training to build a wrapper around the generator for improved sampling. With a perfect discriminator, this wrapped generator samples from the true distribution on the data exactly even when the generator is imperfect. We demonstrate the benefits of the improved generator on multiple benchmark datasets, including CIFAR-10 and CelebA, using DCGAN and WGAN.


Ryan Turner, Jane Hung, Yunus Saatci, Jason Yosinski


ICML 2019

Full Paper

‘Metropolis-Hastings Generative Adversarial Networks’ (PDF)

Uber AI

Previous article Understanding Neural Networks via Feature Visualization: A survey
Next article Exploratory Stage Lighting Design using Visual Objectives
Ryan Turner
Ryan Turner is a senior research scientist with Uber AI Labs.
Jane Hung
Jane Hung is a research scientist with Uber AI Labs.
Yunus Saatci
Yunus Saatci is a senior research scientist with Uber AI Labs.
Jason Yosinski
Jason Yosinski is a founding member of Uber AI Labs and there leads the Deep Collective research group. He is known for contributions to understanding neural network modeling, representations, and training. Prior to Uber, Jason worked on robotics at Caltech, co-founded two web companies, and started a robotics program in Los Angeles middle schools that now serves over 500 students. He completed his PhD working at the Cornell Creative Machines Lab, University of Montreal, JPL, and Google DeepMind. He is a recipient of the NASA Space Technology Research Fellowship, has co-authored over 50 papers and patents, and was VP of ML at Geometric Intelligence, which Uber acquired. His work has been profiled by NPR, the BBC, Wired, The Economist, Science, and the NY Times. In his free time, Jason enjoys cooking, reading, paragliding, and pretending he's an artist.