L1-norm regularization을 통한 SGMM의 state vector 적응

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 709
  • Download : 0
DC FieldValueLanguage
dc.contributor.author구자현ko
dc.contributor.author김영관ko
dc.contributor.author김회린ko
dc.date.accessioned2016-04-14T02:59:22Z-
dc.date.available2016-04-14T02:59:22Z-
dc.date.created2015-11-24-
dc.date.created2015-11-24-
dc.date.issued2015-09-
dc.identifier.citation말소리와 음성과학, v.7, no.3, pp.131 - 138-
dc.identifier.issn2005-8063-
dc.identifier.urihttp://hdl.handle.net/10203/203740-
dc.description.abstractIn this paper, we propose L1-norm regularization for state vector adaptation of subspace Gaussian mixture model (SGMM). When you design a speaker adaptation system with GMM-HMM acoustic model, MAP is the most typical technique to be considered. However, in MAP adaptation procedure, large number of parameters should be updated simultaneously. We can adopt sparse adaptation such as L1-norm regularization or sparse MAP to cope with that, but the performance of sparse adaptation is not good as MAP adaptation. However, SGMM does not suffer a lot from sparse adaptation as GMM-HMM because each Gaussian mean vector in SGMM is defined as a weighted sum of basis vectors, which is much robust to the fluctuation of parameters. Since there are only a few adaptation techniques appropriate for SGMM, our proposed method could be powerful especially when the number of adaptation data is limited. Experimental results show that error reduction rate of the proposed method is better than the result of MAP adaptation of SGMM, even with small adaptation data.-
dc.languageKorean-
dc.publisher한국음성학회-
dc.titleL1-norm regularization을 통한 SGMM의 state vector 적응-
dc.typeArticle-
dc.type.rimsART-
dc.citation.volume7-
dc.citation.issue3-
dc.citation.beginningpage131-
dc.citation.endingpage138-
dc.citation.publicationname말소리와 음성과학-
dc.identifier.doi10.13064/KSSS.2015.7.3.131-
dc.identifier.kciidART002036434-
dc.contributor.localauthor김회린-
dc.subject.keywordAuthorL1-norm regularization-
dc.subject.keywordAuthorspeaker adaptation-
dc.subject.keywordAuthorstate vector adaptation-
dc.subject.keywordAuthorsubspace gaussian mixture model-
dc.subject.keywordAuthorautomatic speech recognition-
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