The geometry of the articulatory region that produces a speech sound

Authors:
Chao Qin;Miguel Á. Carreira-Perpiñán
Affiliations:
EECS, School of Engineering, University of California, Merced;EECS, School of Engineering, University of California, Merced
Venue:
Asilomar'09 Proceedings of the 43rd Asilomar conference on Signals, systems and computers
Year:
2009

Citing 3
Cited 0

Fundamentals of speech recognition

Fundamentals of speech recognition
Mode-Finding for Mixtures of Gaussian Distributions

IEEE Transactions on Pattern Analysis and Machine Intelligence
Data-driven production models for speech processing

Data-driven production models for speech processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

It is known that some speech sounds are produced by more than a single vocal tract shape. Here, we study to what extent individual articulators (e.g. the tongue tip) are constrained by a given acoustic frame. We use parametric and nonparametric methods for articulatory inversion and quantify the error incurred by inversion methods, and the dimensionality and multimodality of the inverse region in articulatory space that corresponds to a speech sound.