File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Conference Paper: ShefCE: A Cantonese-English bilingual speech corpus for pronunciation assessment

TitleShefCE: A Cantonese-English bilingual speech corpus for pronunciation assessment
Authors
KeywordsBilingual parallel speech corpus
Cantonese
English pronunciation assessment
Issue Date2017
PublisherIEEE. The Proceedings' web site is located at http://ieeexplore.ieee.org/xpl/conhome.jsp?punumber=1000002
Citation
2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), New Orleans, LA, 5-9 March 2017 How to Cite?
AbstractThis paper introduces the development of ShefCE: a Cantonese-English bilingual speech corpus from L2 English speakers in Hong Kong. Bilingual parallel recording materials were chosen from TED online lectures. Script selection were carried out according to bilingual consistency (evaluated using a machine translation system) and the distribution balance of phonemes. 31 undergraduate to postgraduate students in Hong Kong aged 20-30 were recruited and recorded a 25-hour speech corpus (12 hours in Cantonese and 13 hours in English). Baseline phoneme/syllable recognition systems were trained on background data with and without the ShefCE training data. The final syllable error rate (SER) for Cantonese is 17.3% and final phoneme error rate (PER) for English is 34.5%. The automatic speech recognition performance on English showed a significant mismatch when applying L1 models on L2 data, suggesting the need for explicit accent adaptation. ShefCE and the corresponding baseline models will be made openly available for academic research.
Persistent Identifierhttp://hdl.handle.net/10722/248686
ISSN

 

DC FieldValueLanguage
dc.contributor.authorNg, RWM-
dc.contributor.authorKwan, ACM-
dc.contributor.authorLee, T-
dc.contributor.authorHain, T-
dc.date.accessioned2017-10-18T08:47:01Z-
dc.date.available2017-10-18T08:47:01Z-
dc.date.issued2017-
dc.identifier.citation2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), New Orleans, LA, 5-9 March 2017-
dc.identifier.issn2379-190X-
dc.identifier.urihttp://hdl.handle.net/10722/248686-
dc.description.abstractThis paper introduces the development of ShefCE: a Cantonese-English bilingual speech corpus from L2 English speakers in Hong Kong. Bilingual parallel recording materials were chosen from TED online lectures. Script selection were carried out according to bilingual consistency (evaluated using a machine translation system) and the distribution balance of phonemes. 31 undergraduate to postgraduate students in Hong Kong aged 20-30 were recruited and recorded a 25-hour speech corpus (12 hours in Cantonese and 13 hours in English). Baseline phoneme/syllable recognition systems were trained on background data with and without the ShefCE training data. The final syllable error rate (SER) for Cantonese is 17.3% and final phoneme error rate (PER) for English is 34.5%. The automatic speech recognition performance on English showed a significant mismatch when applying L1 models on L2 data, suggesting the need for explicit accent adaptation. ShefCE and the corresponding baseline models will be made openly available for academic research.-
dc.languageeng-
dc.publisherIEEE. The Proceedings' web site is located at http://ieeexplore.ieee.org/xpl/conhome.jsp?punumber=1000002-
dc.relation.ispartofIEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)-
dc.rightsIEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). Copyright © IEEE.-
dc.rights©2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.-
dc.subjectBilingual parallel speech corpus-
dc.subjectCantonese-
dc.subjectEnglish pronunciation assessment-
dc.titleShefCE: A Cantonese-English bilingual speech corpus for pronunciation assessment-
dc.typeConference_Paper-
dc.identifier.emailKwan, ACM: cmkwan@hku.hk-
dc.identifier.doi10.1109/ICASSP.2017.7953273-
dc.identifier.hkuros280108-
dc.publisher.placeNew Orleans, LA-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats