1st ETRW on Speech Production Modeling: From Control Strategies to Acoustics 4th Speech Production Seminar: Models and Data Autrans, France May 20-24, 1996

Principal Component Analysis of Cross-Sections of Tongue Shapes in Vowel Production

Maureen Stone (1), Moise H. Goldstein Jr. (2), Yongqing Zhang (3)

(1) University of Maryland School of Medicine, Division of Otolaryngology, Baltimore, MD, USA
(2) The Johns Hopkins University, Department of Electrical and Computer Engineering, Baltimore, MD, USA
(3) The Johns Hopkins University, Department of Mathematical Science, Baltimore, MD, USA

The present study quantified cross-sectional tongue shape for vowels in a single plane using Principal Component Analysis. A single subject repeated eleven English vowels in two consonant contexts, five times. The first two PC's described two waveshapes and accounted for 93% of the variance in the data. They were model-free, and data-derived, thus representing underlying tongue shapes for this subject. The loadings of the eleven vowels on the first two PC's for this subject indicated three distinct shape groups: high vowels, front vowels, and back vowels.

Full Paper

Bibliographic reference.  Stone, Maureen / Goldstein Jr., Moise H. / Zhang, Yongqing (1996): "Principal component analysis of cross-sections of tongue shapes in vowel production", In SPM-1996, 37-40.