Dynamical Variational Autoencoders: A Comprehensive Review

TL;DR

This paper reviews and unifies various dynamical variational autoencoder models that extend VAEs to sequential data, providing a comprehensive overview, reimplementation, and benchmark results on speech analysis-resynthesis.

Contribution

It introduces a general class of dynamical VAEs, standardizes their notation, and benchmarks seven recent models on a speech task.

Findings

Reimplemented seven DVAE models for comparison.

Benchmark results on speech analysis-resynthesis.

Discussion on future directions for DVAE research.

Abstract

Variational autoencoders (VAEs) are powerful deep generative models widely used to represent high-dimensional complex data through a low-dimensional latent space learned in an unsupervised manner. In the original VAE model, the input data vectors are processed independently. Recently, a series of papers have presented different extensions of the VAE to process sequential data, which model not only the latent space but also the temporal dependencies within a sequence of data vectors and corresponding latent vectors, relying on recurrent neural networks or state-space models. In this paper, we perform a literature review of these models. We introduce and discuss a general class of models, called dynamical variational autoencoders (DVAEs), which encompasses a large subset of these temporal VAE extensions. Then, we present in detail seven recently proposed DVAE models, with an aim to homogenize the notations and presentation lines, as well as to relate these models with existing classical temporal models. We have reimplemented those seven DVAE models and present the results of an experimental benchmark conducted on the speech analysis-resynthesis task (the PyTorch code is made publicly available). The paper concludes with a discussion on important issues concerning the DVAE class of models and future research guidelines.