A 3D vocal tract model for articulatory and visual speech synthesis

Olov Engwall

Centre for Speech Technology (CTT)

Department of Speech, Music and Hearing,, KTH, Stockholm



A 3D model of the vocal tract to be used for articulatory and visual synthesis has been developed. The tract is modeled as a smoothly varying tube created of a polygon mesh that can be modified by seven articulatory parameters. The model is to be employed firstly for speech synthesis using e.g. direct calculations of the cross-sectional areas in the tract and secondly as a visual aid for pronounciation training, e.g. of hearing-impaired children.

This article was published in the Proceedings for Fonetik98 in Stockholm, and can be viewed in or printed from the [pdf-version].