Interactive Activation Model of Speech Perception

摘要

The objective of this research was to construct a model of human speech perception, using an interactive activation framework. Computer simulations of several versions of this model were developed. One version accepted real speech as input. A second version took simulated speech input and extracted phonemes, words, and simple phrases. The model was able to account for a variety of phenomena which have been observed in human speech perception; it also suggested behavior which had not previously been observed in human listeners but which was subsequently verified through experimental work.