Sixth International Conference on Spoken Language Processing
(ICSLP 2000)

Beijing, China
October 16-20, 2000

Toward an Acoustic-Articulatory Model of Inter-Speaker Variability

Parham Mokhtari (1), Frantz Clermont (2), Kazuyo Tanaka (1)

(1) Electrotechnical Laboratory, Tsukuba, Ibaraki, Japan
(2) School of Computer Science, Univ. of New South Wales, Canberra, Australia

In this paper we propose a more complete model of inter-speaker variability, which accounts quantitatively for structural differences of the vocal tract (VT) and for learned differences in articulatory setting and in phoneme-specific strategy. This tripartite modelling is applied to a dataset of VT area functions estimated by acoustic-to-articulatory mapping, from formants measured in the steady-states of 5 vowels recorded by 5 adult, male speakers of Japanese. Principal component analyses of each of those physical sources of inter-speaker variation then yield eigenmodes and their variances, which together define an acoustic-articulatory, functional model of inter-speaker variability in vowel production.

Full Paper

Bibliographic reference.  Mokhtari, Parham / Clermont, Frantz / Tanaka, Kazuyo (2000): "Toward an acoustic-articulatory model of inter-speaker variability", In ICSLP-2000, vol.2, 158-161.