Neural Network Modeling of Speech Motor Control Using Physiological Data.

Number 1083
Year 1998
Drawer 20
Entry Date 11/22/1999
Authors Vatikiotis-Bateson, E., Hirayama, M. & Kawato, M.
Contact
Publication ATR Auditory and Visual Perception Research Laboratories. Inuidani, Seika-cho, Soraku-gun, Kyoto 6199—02, Japan.
url http://www.haskins.yale.edu/Reprints/HL1083.pdf
Abstract [Introduction] In this paper, we focus on the dynamical relation between observed elements of neuromuscular activity, indicative of neuromotor commands, and the ensuing articulatory movement behavior during reiterate speech production. Although progress has been made modeling the articulatory kinematics from which inferences can be made about the underlying dynamics of the musculoskeletal system (for review, Saltzman & Munhall, 1989), the mapping between actual neuromuscular and articulatory events- in which the system’s dynamical properties could be directly assess- still remains a mystery. Ultimately, we hope this computational approach to speech production can be used to model speech perception and recognition, and to elucidate structural and functional aspects of the central nervous system’s orchestration of motor behavior.
Notes

Search Publications