14th International Congress of Phonetic Sciences (ICPhS-14) San Francisco, CA, USA August 1-7, 1999

Constructing Speaker-Specific Articulatory Vocal Tract Models for Testing Speech Motor Control Hypotheses

Alfonso Nieto-Castanon (1), Frank H. Guenther (1,2)

(1) Department of Cognitive and Neural Systems, Boston University, USA
(2) Research Laboratory of Electronics, MIT, USA

Inter-speaker differences in the articulation of speech sounds are common, presumably due in part to differences in vocal tract morphology. These differences can confound the process of using articulator movement data to evaluate theories of speech motor control. This abstract describes software that allows relatively fast construction of speaker-specific articulatory vocal tract models from midsagittal vocal tract profiles, acquired with Magnetic Resonance Imaging (MRI) while speakers produce a list of approximately 30 phonemes. The resulting vocal tract models will be used in concert with experimental data to test models of speech motor control.

Full Paper

Bibliographic reference.  Nieto-Castanon, Alfonso / Guenther, Frank H. (1999): "Constructing speaker-specific articulatory vocal tract models for testing speech motor control hypotheses", In ICPhS-14, 2271-2274.