DC Field | Value | Language |
---|---|---|
dc.contributor.author | Suh, Young Joo | ko |
dc.contributor.author | Kim, HoiRin | ko |
dc.date.accessioned | 2016-04-25T05:29:55Z | - |
dc.date.available | 2016-04-25T05:29:55Z | - |
dc.date.created | 2013-02-13 | - |
dc.date.created | 2013-02-13 | - |
dc.date.issued | 2013-04 | - |
dc.identifier.citation | JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA, v.133, no.4, pp.307 - 313 | - |
dc.identifier.issn | 0001-4966 | - |
dc.identifier.uri | http://hdl.handle.net/10203/206178 | - |
dc.description.abstract | Support vector machines (SVMs) have been proved to be an effective approach to speaker verification. An appropriate selection of the kernel function is a key issue in SVM-based classification. In this letter, a new SVM-based speaker verification method utilizing weighted kernels in the Gaussian mixture model supervector space is proposed. The weighted kernels are derived by using the discriminative training approach, which minimizes speaker verification errors. Experiments performed on the NIST 2008 speaker recognition evaluation task showed that the proposed approach provides substantially improved performance over the baseline kernel-based method. (C) 2013 Acoustical Society of America | - |
dc.language | English | - |
dc.publisher | ACOUSTICAL SOC AMER AMER INST PHYSICS | - |
dc.subject | SPEECH RECOGNITION | - |
dc.title | Minimum classification error-based weighted support vector machine kernels for speaker verification | - |
dc.type | Article | - |
dc.identifier.wosid | 000318555300014 | - |
dc.identifier.scopusid | 2-s2.0-84876141077 | - |
dc.type.rims | ART | - |
dc.citation.volume | 133 | - |
dc.citation.issue | 4 | - |
dc.citation.beginningpage | 307 | - |
dc.citation.endingpage | 313 | - |
dc.citation.publicationname | JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA | - |
dc.identifier.doi | 10.1121/1.4794350 | - |
dc.contributor.localauthor | Kim, HoiRin | - |
dc.type.journalArticle | Article | - |
dc.subject.keywordPlus | SPEECH RECOGNITION | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.