Speech Technologies: Language Variation

Research output: Chapter in Book/Report/Conference proceedingOther chapter contribution

Abstract

Spoken language technologies, such as automatic speech recognition, must often deal with speech from many speakers and in a wide variety of situations. Variation in speech usually creates serious difficulties for such systems. This chapter looks at the many sources of variation in speech: within a single speaker, across speakers and languages, and in external factors.
Original languageEnglish
Title of host publicationEncyclopedia of Language and Linguistics
EditorsKeith Brown
PublisherElsevier
Pages56-61
Number of pages6
Edition2nd
ISBN (Print)978-0-08-044854-1
DOIs
Publication statusPublished - Mar 2006

Keywords

  • automatic speech recognition
  • co-articulation
  • speech signal processing
  • speech synthesis
  • text-to-speech synthesis
  • voice transformation

Fingerprint Dive into the research topics of 'Speech Technologies: Language Variation'. Together they form a unique fingerprint.

Cite this