InterSpeech 2021

A Benchmark of Dynamical Variational Autoencoders applied to Speech Spectrogram Modeling
(Oral presentation)

Xiaoyu Bie (LJK (UMR 5224), France), Laurent Girin (GIPSA-lab (UMR 5216), France), Simon Leglaive (IETR (UMR 6164), France), Thomas Hueber (GIPSA-lab (UMR 5216), France), Xavier Alameda-Pineda (LJK (UMR 5224), France)
The Variational Autoencoder (VAE) is a powerful deep generative model that is now extensively used to represent high-dimensional complex data via a low-dimensional latent space learned in an unsupervised manner. In the original VAE model, input data vectors are processed independently. In recent years, a series of papers have presented different extensions of the VAE to process sequential data, that not only model the latent space, but also model the temporal dependencies within a sequence of data vectors and corresponding latent vectors, relying on recurrent neural networks. We recently performed a comprehensive review of those models and unified them into a general class called Dynamical Variational Autoencoders (DVAEs). In the present paper, we present the results of an experimental benchmark comparing six of those DVAE models on the speech analysis-resynthesis task, as an illustration of the high potential of DVAEs for speech modeling.