Improving VAE generations of multimodal data through data-dependent conditional priors

TL;DR

This paper introduces CP-VAE, a novel variational autoencoder that learns data-dependent conditional priors to improve generation from individual modalities of multimodal data, addressing a key limitation of traditional VAEs.

Contribution

The paper proposes a new formulation of VAE with conditional priors that differentiate mixture components, enabling modality-specific data generation in an unsupervised manner.

Findings

Enhanced generation from individual data modalities.

Outperforms baseline models in generative tasks.

Effectively models multimodal data structure.

Abstract

One of the major shortcomings of variational autoencoders is the inability to produce generations from the individual modalities of data originating from mixture distributions. This is primarily due to the use of a simple isotropic Gaussian as the prior for the latent code in the ancestral sampling procedure for the data generations. We propose a novel formulation of variational autoencoders, conditional prior VAE (CP-VAE), which learns to differentiate between the individual mixture components and therefore allows for generations from the distributional data clusters. We assume a two-level generative process with a continuous (Gaussian) latent variable sampled conditionally on a discrete (categorical) latent component. The new variational objective naturally couples the learning of the posterior and prior conditionals, and the learning of the latent categories encoding the multimodality of the original data in an unsupervised manner. The data-dependent conditional priors are then used to sample the continuous latent code when generating new samples from the individual mixture components corresponding to the multimodal structure of the original data. Our experimental results illustrate the generative performance of our new model comparing to multiple baselines.