Variational Diffusion Auto-encoder: Deep Latent Variable Model with Unconditional Diffusion Prior
share
Variational auto-encoders (VAEs) are one of the most popular approaches to deep generative modeling. Despite their success, images generated by VAEs are known to suffer from blurriness, due to a highly unrealistic modeling assumption that the conditional data distribution p(x | z) can be approximated as an isotropic Gaussian. In this work we introduce a principled approach to modeling the conditional data distribution p(x | z) by incorporating a diffusion model. We show that it is possible to create a VAE-like deep latent variable model without making the Gaussian assumption on p(x | z) or even training a decoder network. A trained encoder and an unconditional diffusion model can be combined via Bayes' rule for score functions to obtain an expressive model for p(x | z). Our approach avoids making strong assumptions on the parametric form of p(x | z), and thus allows to significantly improve the performance of VAEs.
READ FULL TEXT