<?xml version="1.0" standalone="yes"?>
<Paper uid="N04-3003">
  <Title>Limited-Domain Speech-to-Speech Translation between English and Pashto</Title>
  <Section position="7" start_page="0" end_page="0" type="concl">
    <SectionTitle>
6 Summary
</SectionTitle>
    <Paragraph position="0"> We have described a prototype of a spoken language translation system for use by English-speaking medical personnel treating Pashto-speaking patients. The system, which is targeted to a handheld device, is being developed with extremely little Pashto-language data of any kind. It builds on medium-vocabulary speaker-independent HMM-based speech recognition, rule-based translation and supporting methods, and concatenative synthesis. While the system is certainly still under development, it has provided a reasonable proof of concept in informal trials.</Paragraph>
  </Section>
class="xml-element"></Paper>