Metropolis-Hastings Generative Adversarial Networks
This addresses sampling quality issues in generative modeling for AI applications, but it is incremental as it builds on existing GAN frameworks.
The paper tackles the problem of imperfect sampling in GANs by introducing MH-GAN, which uses the discriminator to wrap the generator for improved sampling, achieving exact sampling from the true distribution with a perfect discriminator and demonstrating benefits on datasets like CIFAR-10 and CelebA.
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 standard GANs which draw samples from the distribution defined only 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 the DCGAN, WGAN, and progressive GAN.