<?xml version="1.0" standalone="yes"?>
<Paper uid="P04-1008">
  <Title>Statistical Modeling for Unit Selection in Speech Synthesis</Title>
  <Section position="7" start_page="0" end_page="0" type="concl">
    <SectionTitle>
6 Conclusion
</SectionTitle>
    <Paragraph position="0"> We introduced a statistical modeling approach to unit selection in speech synthesis. This approach is likely to lead to more accurate unit selection systems based on principled learning algorithms and techniques that radically depart from the heuristic methods used in the traditional systems. Our preliminary experiments using a training corpus generated by the AT&amp;T Natural Voices Product demonstrates that statistical modeling techniques can be used to build a high-quality unit selection system.</Paragraph>
    <Paragraph position="1"> It also shows other important benefits of this approach: a substantial increase of efficiency and a greater modularity and flexibility.</Paragraph>
  </Section>
class="xml-element"></Paper>