Third ESCA/COCOSDA Workshop on Speech Synthesis

November 26-29, 1998
Jenolan Caves House, Blue Mountains, NSW, Australia

A Three-dimensional Linear Articulatory Model Based on MRI Data

Pierre Badin (1), Gérard Bailly (1), M. Raybaudi (2), C. Segebarth (2)

(1) Institut de la Communication Parlée, UPRESA CNRS 5009, INPG - Univ. Stendhal, Grenoble, France
(2) Unité INSERM U438, Grenoble, France

Based on a set of 3D vocal tract images obtained by MRI, a 3D statistical articulatory model has been built using guided Principal Component Analysis. It constitutes an extension to the lateral dimension of the mid-sagittal model previously developed from a radiofilm recorded on the same subject. The parameters of the 2D model have been found to be good predictors of the 3D shapes, for most configurations. A first evaluation of the model in terms of area functions and formants is presented.

Full Paper

Bibliographic reference.  Badin, Pierre / Bailly, Gérard / Raybaudi, M. / Segebarth, C. (1998): "A Three-dimensional Linear Articulatory Model Based on MRI Data", In SSW3-1998, 249-254.