Modulation spectrum-constrained trajectory error training for mixture density network-based speech synthesis

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 171
  • Download : 0
DC FieldValueLanguage
dc.contributor.authorPark, Sangjunko
dc.contributor.authorHahn, Minsooko
dc.date.accessioned2018-12-20T06:52:44Z-
dc.date.available2018-12-20T06:52:44Z-
dc.date.created2018-11-30-
dc.date.created2018-11-30-
dc.date.issued2018-09-
dc.identifier.citationJOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA, v.144, no.3, pp.EL151 - EL157-
dc.identifier.issn0001-4966-
dc.identifier.urihttp://hdl.handle.net/10203/248313-
dc.description.abstractIn statistical parametric speech synthesis, a mixture density network is employed to address the limitations of a linear output layer such as pre-computed fixed variances and the unimodal assumption. However, it also has a defect, i.e., it cannot deploy a static-dynamic constraint needed in the training phase for high-quality speech synthesis. To cope with this problem, this paper proposes a training algorithm based on the minimum trajectory error for a mixture density network. And a modulation spectrum-constrained loss function is also proposed to alleviate the over-smoothing effect. The experimental results confirm meaningful improvement both in objective and subjective performance measures.-
dc.languageEnglish-
dc.publisherACOUSTICAL SOC AMER AMER INST PHYSICS-
dc.titleModulation spectrum-constrained trajectory error training for mixture density network-based speech synthesis-
dc.typeArticle-
dc.identifier.wosid000457802200001-
dc.identifier.scopusid2-s2.0-85052990891-
dc.type.rimsART-
dc.citation.volume144-
dc.citation.issue3-
dc.citation.beginningpageEL151-
dc.citation.endingpageEL157-
dc.citation.publicationnameJOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA-
dc.identifier.doi10.1121/1.5052206-
dc.contributor.localauthorHahn, Minsoo-
dc.description.isOpenAccessN-
dc.type.journalArticleArticle-
Appears in Collection
EE-Journal Papers(저널논문)
Files in This Item
There are no files associated with this item.

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0