<?xml version="1.0" standalone="yes"?>
<Paper uid="H89-2046">
  <Title>THE AUDITORY PROCESSING AND RECOGNITION OF SPEECH</Title>
  <Section position="3" start_page="0" end_page="0" type="intro">
    <SectionTitle>
INTRODUCTION
</SectionTitle>
    <Paragraph position="0"> Systems for the automatic recognition of speech have in recent years derived many ideas and strategies from observations of the structure and processing modes of the nervous system, and specifically of the mammalian auditory system. Examples range from the adoption of cochlear-like processing as front-end analysis stages, to the use of artificial neural networks as adaptive pattern recognizers. For the last five years, we have been studying the functions and algorithms that facilitate the remarkable abilities of the auditory system to analyze, recognize, and localize complex sounds such as speech, music, and other environmental sounds.</Paragraph>
    <Paragraph position="1"> We have developed and used biophysical and computational models of the peripheral cochlear stages, of the intermediate central neural networks that extract various feature representations of the acoustic stimulus (e.g., as in speech phonemes), and of networks for the recognition of temporally ordered sequences (such as words and sentences). These models have been described in detail in \[1,2,3,4,5\]. Here, we shall outline a few of our recent investigations and the results that we have obtained.</Paragraph>
  </Section>
class="xml-element"></Paper>