Diffusion Priors In Variational Autoencoders

By Antoine Wehenkel, Gilles Louppe. 2021

34 citations recorded by the source corpus.

Among likelihood-based approaches for deep generative modelling, variational autoencoders (VAEs) offer scalable amortized posterior inference and fast sampling. However, VAEs are also more and more outperformed by competing models such as normalizing flows (NFs), deep-energy models, or the new denoising diffusion probabilistic models (DDPMs). In this preliminary work, we improve VAEs by demonstrating how DDPMs can be used for modelling the prior distribution of the latent variables. The diffusion prior model improves upon Gaussian priors of classical VAEs and is competitive with NF-based priors. Finally, we hypothesize that hierarchical VAEs could similarly benefit from the enhanced capacity of diffusion priors.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Statistical and Nonlinear Physics, Generative Models, Probabilistic & Bayesian Methods

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Statistical and Nonlinear Physics

In Generative Models

In Probabilistic & Bayesian Methods

Citations in the corpus

Score-based Generative Modeling in Latent Space (2021, background): “However, [10, 69, 70] consider the discretized diffusion objective [1], while we build on the continuous time framework.”
Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow (2022, unknown): “[68, 89, 81] learns a denoising diffusion model in the latent spaces and combines it with variational auto-encoders.”
Diffusion Models: A Comprehensive Survey of Methods and Applications (2022, methodology): “Alternatively, Wehenkel and Louppe [224] address this problem by designing the VLB of diffusion model combined with reconstruction loss of VAE.”