ISCA Archive Interspeech 2021
ISCA Archive Interspeech 2021

Learning Robust Speech Representation with an Articulatory-Regularized Variational Autoencoder

Marc-Antoine Georges, Laurent Girin, Jean-Luc Schwartz, Thomas Hueber

It is increasingly considered that human speech perception and production both rely on articulatory representations. In this paper, we investigate whether this type of representation could improve the performances of a deep generative model (here a variational autoencoder) trained to encode and decode acoustic speech features. First we develop an articulatory model able to associate articulatory parameters describing the jaw, tongue, lips and velum configurations with vocal tract shapes and spectral features. Then we incorporate these articulatory parameters into a variational autoencoder applied on spectral features by using a regularization technique that constrains part of the latent space to represent articulatory trajectories. We show that this articulatory constraint improves model training by decreasing time to convergence and reconstruction loss at convergence, and yields better performance in a speech denoising task.


doi: 10.21437/Interspeech.2021-1604

Cite as: Georges, M.-A., Girin, L., Schwartz, J.-L., Hueber, T. (2021) Learning Robust Speech Representation with an Articulatory-Regularized Variational Autoencoder. Proc. Interspeech 2021, 3345-3349, doi: 10.21437/Interspeech.2021-1604

@inproceedings{georges21_interspeech,
  author={Marc-Antoine Georges and Laurent Girin and Jean-Luc Schwartz and Thomas Hueber},
  title={{Learning Robust Speech Representation with an Articulatory-Regularized Variational Autoencoder}},
  year=2021,
  booktitle={Proc. Interspeech 2021},
  pages={3345--3349},
  doi={10.21437/Interspeech.2021-1604}
}