ResearchSpace

HMM Adaptation for child speech synthesis

Show simple item record

dc.contributor.author Govender, Avashna
dc.contributor.author De Wet, Febe
dc.contributor.author Tapamo, Jules-Raymond
dc.date.accessioned 2017-09-04T12:34:30Z
dc.date.available 2017-09-04T12:34:30Z
dc.date.issued 2015-09
dc.identifier.citation Govender, A., De Wet, F. and Tapamo, J. 2015. HMM adaptation for child speech synthesis. 16th Annual Conference of the International Speech Communication Association (Interspeech 2015), 6 - 10 September 2015, Dresden, Germany, pp 1640-1644. en_US
dc.identifier.uri http://interspeech2015.org/
dc.identifier.uri http://hdl.handle.net/10204/9529
dc.description Copyright: 2015 Technische Universität Berlin. Due to copyright restrictions, the attached PDF file only contains the abstract of the full text item. For access to the full text item, kindly consult the publisher's website. en_US
dc.description.abstract Hidden Markov Model (HMM)-based synthesis in combination with speaker adaptation has proven to be an approach that is well-suited for child speech synthesis. This paper describes the development and evaluation of different HMM-based child speech synthesis systems. The aim is to determine the most suitable combination of initial model and speaker adaptation techniques to synthesize child speech. The results of the study indicate that gender-independent initial models perform better than gender-dependent initial models and Constrained Structural Maximum a Posteriori Linear Regression (CSMAPLR) followed by maximum a posteriori (MAP) is the speaker adaptation technique combination that yields the most natural and intelligible synthesized child speech. en_US
dc.language.iso en en_US
dc.publisher Technische Universität Berlin en_US
dc.relation.ispartofseries Worklist;15702
dc.subject HMM-based synthesis en_US
dc.subject Hidden Markov Model en_US
dc.subject Speaker adaptation en_US
dc.subject Average-voice-based synthesis en_US
dc.subject Child speech synthesis en_US
dc.title HMM Adaptation for child speech synthesis en_US
dc.type Conference Presentation en_US
dc.identifier.apacitation Govender, A., De Wet, F., & Tapamo, J. (2015). HMM Adaptation for child speech synthesis. Technische Universität Berlin. http://hdl.handle.net/10204/9529 en_ZA
dc.identifier.chicagocitation Govender, Avashna, Febe De Wet, and Jules-Raymond Tapamo. "HMM Adaptation for child speech synthesis." (2015): http://hdl.handle.net/10204/9529 en_ZA
dc.identifier.vancouvercitation Govender A, De Wet F, Tapamo J, HMM Adaptation for child speech synthesis; Technische Universität Berlin; 2015. http://hdl.handle.net/10204/9529 . en_ZA
dc.identifier.ris TY - Conference Presentation AU - Govender, Avashna AU - De Wet, Febe AU - Tapamo, Jules-Raymond AB - Hidden Markov Model (HMM)-based synthesis in combination with speaker adaptation has proven to be an approach that is well-suited for child speech synthesis. This paper describes the development and evaluation of different HMM-based child speech synthesis systems. The aim is to determine the most suitable combination of initial model and speaker adaptation techniques to synthesize child speech. The results of the study indicate that gender-independent initial models perform better than gender-dependent initial models and Constrained Structural Maximum a Posteriori Linear Regression (CSMAPLR) followed by maximum a posteriori (MAP) is the speaker adaptation technique combination that yields the most natural and intelligible synthesized child speech. DA - 2015-09 DB - ResearchSpace DP - CSIR KW - HMM-based synthesis KW - Hidden Markov Model KW - Speaker adaptation KW - Average-voice-based synthesis KW - Child speech synthesis LK - https://researchspace.csir.co.za PY - 2015 T1 - HMM Adaptation for child speech synthesis TI - HMM Adaptation for child speech synthesis UR - http://hdl.handle.net/10204/9529 ER - en_ZA


Files in this item

Thumbnail
Name: Govender_15702_20 ...
Size: 5.083Kb
Format: PDF
View/Open

This item appears in the following Collection(s)

Show simple item record