<?xml version="1.0" standalone="yes"?>
<Paper uid="H94-1078">
  <Title>TECHNIQUES TO ACHIEVE AN ACCURATE REAL-TIME LARGE- VOCABULARY SPEECH RECOGNITION SYSTEM</Title>
  <Section position="7" start_page="397" end_page="397" type="concl">
    <SectionTitle>
6. CONCLUSIONS
</SectionTitle>
    <Paragraph position="0"> Tree-based techniques, combined with appropriate modeling alternatives, can achieve real-time performance at about 30% error rate for ARPA's 20,000-word Wall Street Journal task. We have shown techniques that reduce the computational complexity of more accurate but slower modeling alternatives so that they are near the speed necessary for real-time performance in a multipass search. Our near-future goal is to combine these two technologies so that real-time, high-accuracy large-vocabulary speech recognition can be achieved.</Paragraph>
  </Section>
class="xml-element"></Paper>