Wiki Cascaded-Gaussian Mixture Regression¶

What is C-GMR?¶

Cascaded Gaussian Mixture Regression or C-GMR is a general framework for adapting a GMR (Gaussian Mixture Regression) trained on a large dataset of input-output joint observations, using a limited set of input-only observations. It was originaly developed in the context of speech processing fpr adapting an acoustic-articulatory inversion GMM trained on a reference speaker, to a new speaker, given a small amount of audio-only observations. In particular, C-GMR framework includes the "Integrated C-GMR" model (IC-GMR) which combines 2 consecutive GMR in a single probabilistic model. For this model, we derived both the exact EM-based training algorithm and inference equation.

How-to-cite¶

T. Hueber, L. Girin, X. Alameda-Pineda, and G. Bailly, « Speaker-adaptive acoustic-articulatory inversion using cascaded Gaussian mixture regression, », IEEE Transactions on Audio, Speech and Language Processing, July 2015, in press.

Source code¶

The Matlab source code for training and using the Integrated C-GMR and the JointGMR can be downloaded from the C-GMR git repository https://git.gipsa-lab.grenoble-inp.fr/cgmr.git

Contributors¶

• Dr. Thomas Hueber, CNRS researcher, GIPSA-lab (Grenoble, France)
• Dr. Laurent Girin, Professor at Grenoble-INP, GIPSA-lab/INRIA (Grenoble, France)
• Dr. Xavier Alameda-Pineda, Researcher at UNITN (Trento, Italy)
• Dr. Gérard Bailly, CNRS researcher, GIPSA-lab (Grenoble, France)

Contact¶

thomas.hueber@gipsa-lab.fr