ResearchSpace

Automatic error detection in alignments for speech synthesis

Show simple item record

dc.contributor.author Barnard, E
dc.contributor.author Davel, M
dc.date.accessioned 2007-07-27T10:00:56Z
dc.date.available 2007-07-27T10:00:56Z
dc.date.issued 2006-11
dc.identifier.citation Barnard, E and Davel, M. 2006. Automatic error detection in alignments for speech synthesis. 17th Annual Symposium of the Pattern Recognition Association of South Africa, Parys, South Africa, 29 Nov - 1 Dec 2006, pp 4 en
dc.identifier.isbn 978-0-620-37384-5
dc.identifier.uri http://hdl.handle.net/10204/1044
dc.description.abstract The phonetic segmentation of recorded speech is a crucial factor in the quality of concatenative systems for speech synthesis. The authors describe a likelihood-based error detection process that can be used to flag possible errors in such a segmentation, with a view towards manual correction. It is shown that this process can be used to assist in the creation of high-accuracy segmentations. In particular, for an isiZulu corpus used in the creation of a unit-selection synthesizer, almost half of the errors that existed in a manual segmentation were detected by this process, while flagging less than a quarter of all segments. Different phoneme classes are handled with differing amounts of success, with vowels being the most troublesome en
dc.language.iso en en
dc.subject Speech synthesizer en
dc.subject Phonetic segmentation en
dc.subject Error detection en
dc.subject Unit-selection synthesizer en
dc.title Automatic error detection in alignments for speech synthesis en
dc.type Conference Presentation en


Files in this item

This item appears in the following Collection(s)

Show simple item record