Learning Priors for Adversarial Autoencoders

TL;DR

This paper introduces a method to learn data-driven priors for adversarial autoencoders, improving image quality and disentanglement, and enabling cross-domain translation in text-to-image synthesis.

Contribution

It proposes code generators to learn priors from data, enhancing the expressiveness and performance of adversarial autoencoders.

Findings

Better image quality in generated outputs

Improved disentangled representations

Effective cross-domain translation

Abstract

Most deep latent factor models choose simple priors for simplicity, tractability or not knowing what prior to use. Recent studies show that the choice of the prior may have a profound effect on the expressiveness of the model,especially when its generative network has limited capacity. In this paper, we propose to learn a proper prior from data for adversarial autoencoders(AAEs). We introduce the notion of code generators to transform manually selected simple priors into ones that can better characterize the data distribution. Experimental results show that the proposed model can generate better image quality and learn better disentangled representations than AAEs in both supervised and unsupervised settings. Lastly, we present its ability to do cross-domain translation in a text-to-image synthesis task.