Adversarial Variational Bayes: Unifying Variational Autoencoders and Generative Adversarial Networks

TL;DR

This paper introduces Adversarial Variational Bayes (AVB), a novel training method for VAEs that employs an auxiliary discriminator to enable highly expressive inference models, bridging VAEs and GANs with strong theoretical foundations.

Contribution

The paper presents AVB, a new approach that unifies VAEs and GANs, providing a theoretically justified method for training VAEs with arbitrarily expressive inference models.

Findings

Exact maximum-likelihood estimation in the nonparametric limit

AVB retains advantages of standard VAEs

Easy to implement and theoretically justified

Abstract

Variational Autoencoders (VAEs) are expressive latent variable models that can be used to learn complex probability distributions from training data. However, the quality of the resulting model crucially relies on the expressiveness of the inference model. We introduce Adversarial Variational Bayes (AVB), a technique for training Variational Autoencoders with arbitrarily expressive inference models. We achieve this by introducing an auxiliary discriminative network that allows to rephrase the maximum-likelihood-problem as a two-player game, hence establishing a principled connection between VAEs and Generative Adversarial Networks (GANs). We show that in the nonparametric limit our method yields an exact maximum-likelihood assignment for the parameters of the generative model, as well as the exact posterior distribution over the latent variables given an observation. Contrary to competing approaches which combine VAEs with GANs, our approach has a clear theoretical justification, retains most advantages of standard Variational Autoencoders and is easy to implement.