Communication between Speech Production and Perception - Observation and Simulation

摘要

Advancement of human-machine interface requires us to investigate human mechanisms and implement them in a machine. This study focuses on communication between speech production and speech perception within human brain and its realization in a cyber system. A physiological research based on electromyographic signals (Honda, 1996) suggested that speech communication system in human brain might be a topologically mapping between speech production and speech perception according to an analogous topology between motor and sensory representations. Following this hypothesis, we examined the relation between the muscle force and tongue movement using a physiological articulatory model. The results indicated that each muscle drives the tongue towards an equilibrium position (EP) corresponding to the magnitude of the muscle forces. Accordingly, an invariant mapping, the EP map. established the connection of spatial locations and muscle forces. Two types of forces, an EP-map based static force and a workspace-based dynamic force, were taken into account of model computation. The simulation showed that the mapping from the muscle force space and articulatory space was unique, while there were multiple choices for muscle force to achieve a given target. The motor-to-kinematic-to-acoustic flow in the model simulation showed the compatibility among the three spaces.