Adaptation of a Gaussian Mixture Regressor to a New Input Distribution: Extending the C-GMR Framework

Laurent Girin (1, 2) , Thomas Hueber (1) , Xavier Alameda-Pineda (3, 2)

1 GIPSA-CRISSP - GIPSA - Cognitive Robotics, Interactive Systems, & Speech Processing
2 PERCEPTION - Interpretation and Modelling of Images and Videos
3 UNITN - Università degli Studi di Trento = University of Trento

This paper addresses the problem of the adaptation of a Gaussian Mixture Regression (GMR) to a new input distribution, using a limited amount of input-only examples. We propose a new model for GMR adaptation, called Joint GMR (J-GMR), that extends the previously published framework of Cascaded GMR (C-GMR). We provide an exact EM training algorithm for the J-GMR. We discuss the merits of the J-GMR with respect to the C-GMR and illustrate its performance with experiments on speech acoustic-to-articulatory inversion.

Mots clés: Gaussian Mixture Regression; GMM; Adaptation; EM algorithm

Domaines: Informatique [cs]; Traitement du signal et de l'image [eess.SP]

Laurent Girin, Thomas Hueber, Xavier Alameda-Pineda. Adaptation of a Gaussian Mixture Regressor to a New Input Distribution: Extending the C-GMR Framework. LVA/ICA 2017 - 13th International Conference on Latent Variable Analysis and Signal Separation, Feb 2017, Grenoble, France. pp.459-468, ⟨10.1007/978-3-319-53547-0_43⟩. ⟨hal-01646098⟩