A unified maximum likelihood approach to acoustic mismatch compensation : Application to noisy lombard speech recognition
Identifieur interne : 00C542 ( Main/Merge ); précédent : 00C541; suivant : 00C543A unified maximum likelihood approach to acoustic mismatch compensation : Application to noisy lombard speech recognition
Auteurs : M. Afify [France] ; Y. Gong [États-Unis] ; Jean-Paul Haton [France]Source :
Descripteurs français
- Pascal (Inist)
English descriptors
- KwdEn :
Abstract
In the context of continuous density hidden Markov model (CDHMM) we present a unified maximum likelihood (ML) approach to acoustic mismatch compensation. This is achieved by introducing additive Gaussian biases at the state level in both the mel cepstral and linear spectral domains. Flexible modelling of different mismatch effects can be obtained through appropriate bias tying. A Maximum likelihood approach for joint estimation of both mel cepstral and linear spectral biases from the observed mismatched speech given only one set of clean speech models is presented, where the obtained bias estimates are used for the compensation of clean speech models during decoding. The proposed approach is applied to the recognition of noisy Lombard speech, and significant improvement in the word recognition rate is achieved.
Links toward previous steps (curation, corpus...)
- to stream PascalFrancis, to step Corpus: 000C14
- to stream PascalFrancis, to step Curation: 000C60
- to stream PascalFrancis, to step Checkpoint: 000C45
Links to Exploration step
Pascal:98-0082849Le document en format XML
<record><TEI><teiHeader><fileDesc><titleStmt><title xml:lang="en" level="a">A unified maximum likelihood approach to acoustic mismatch compensation : Application to noisy lombard speech recognition</title><author><name sortKey="Afify, M" sort="Afify, M" uniqKey="Afify M" first="M." last="Afify">M. Afify</name><affiliation wicri:level="3"><inist:fA14 i1="01"><s1>CRIN/CNRS-INRIA-Lorraine, B.P. 239</s1><s2>54506 Vandeouvre, Nancy</s2><s3>FRA</s3><sZ>1 aut.</sZ><sZ>3 aut.</sZ></inist:fA14><country>France</country><placeName><region type="region">Grand Est</region><region type="old region">Lorraine (région)</region><settlement type="city">Nancy</settlement></placeName></affiliation></author><author><name sortKey="Gong, Y" sort="Gong, Y" uniqKey="Gong Y" first="Y." last="Gong">Y. Gong</name><affiliation wicri:level="1"><inist:fA14 i1="02"><s1>Speech Research, Personal Systems Laboratory, Texas Instruments, P.O. BOX 655303 MS 8374</s1><s2>Dallas TX 75265</s2><s3>USA</s3><sZ>2 aut.</sZ></inist:fA14><country>États-Unis</country><wicri:noRegion>Dallas TX 75265</wicri:noRegion></affiliation></author><author><name sortKey="Haton, J P" sort="Haton, J P" uniqKey="Haton J" first="J.-P." last="Haton">Jean-Paul Haton</name><affiliation wicri:level="3"><inist:fA14 i1="01"><s1>CRIN/CNRS-INRIA-Lorraine, B.P. 239</s1><s2>54506 Vandeouvre, Nancy</s2><s3>FRA</s3><sZ>1 aut.</sZ><sZ>3 aut.</sZ></inist:fA14><country>France</country><placeName><region type="region">Grand Est</region><region type="old region">Lorraine (région)</region><settlement type="city">Nancy</settlement></placeName><placeName><settlement type="city">Nancy</settlement><region type="region" nuts="2">Grand Est</region><region type="region" nuts="2">Lorraine (région)</region></placeName><orgName type="laboratoire" n="5">Laboratoire lorrain de recherche en informatique et ses applications</orgName><orgName type="university">Université de Lorraine</orgName><orgName type="institution">Centre national de la recherche scientifique</orgName><orgName type="institution">Institut national de recherche en informatique et en automatique</orgName></affiliation></author></titleStmt><publicationStmt><idno type="wicri:source">INIST</idno><idno type="inist">98-0082849</idno><date when="1997">1997</date><idno type="stanalyst">PASCAL 98-0082849 INIST</idno><idno type="RBID">Pascal:98-0082849</idno><idno type="wicri:Area/PascalFrancis/Corpus">000C14</idno><idno type="wicri:Area/PascalFrancis/Curation">000C60</idno><idno type="wicri:Area/PascalFrancis/Checkpoint">000C45</idno><idno type="wicri:explorRef" wicri:stream="PascalFrancis" wicri:step="Checkpoint">000C45</idno><idno type="wicri:Area/Main/Merge">00C542</idno></publicationStmt><sourceDesc><biblStruct><analytic><title xml:lang="en" level="a">A unified maximum likelihood approach to acoustic mismatch compensation : Application to noisy lombard speech recognition</title><author><name sortKey="Afify, M" sort="Afify, M" uniqKey="Afify M" first="M." last="Afify">M. Afify</name><affiliation wicri:level="3"><inist:fA14 i1="01"><s1>CRIN/CNRS-INRIA-Lorraine, B.P. 239</s1><s2>54506 Vandeouvre, Nancy</s2><s3>FRA</s3><sZ>1 aut.</sZ><sZ>3 aut.</sZ></inist:fA14><country>France</country><placeName><region type="region">Grand Est</region><region type="old region">Lorraine (région)</region><settlement type="city">Nancy</settlement></placeName></affiliation></author><author><name sortKey="Gong, Y" sort="Gong, Y" uniqKey="Gong Y" first="Y." last="Gong">Y. Gong</name><affiliation wicri:level="1"><inist:fA14 i1="02"><s1>Speech Research, Personal Systems Laboratory, Texas Instruments, P.O. BOX 655303 MS 8374</s1><s2>Dallas TX 75265</s2><s3>USA</s3><sZ>2 aut.</sZ></inist:fA14><country>États-Unis</country><wicri:noRegion>Dallas TX 75265</wicri:noRegion></affiliation></author><author><name sortKey="Haton, J P" sort="Haton, J P" uniqKey="Haton J" first="J.-P." last="Haton">Jean-Paul Haton</name><affiliation wicri:level="3"><inist:fA14 i1="01"><s1>CRIN/CNRS-INRIA-Lorraine, B.P. 239</s1><s2>54506 Vandeouvre, Nancy</s2><s3>FRA</s3><sZ>1 aut.</sZ><sZ>3 aut.</sZ></inist:fA14><country>France</country><placeName><region type="region">Grand Est</region><region type="old region">Lorraine (région)</region><settlement type="city">Nancy</settlement></placeName><placeName><settlement type="city">Nancy</settlement><region type="region" nuts="2">Grand Est</region><region type="region" nuts="2">Lorraine (région)</region></placeName><orgName type="laboratoire" n="5">Laboratoire lorrain de recherche en informatique et ses applications</orgName><orgName type="university">Université de Lorraine</orgName><orgName type="institution">Centre national de la recherche scientifique</orgName><orgName type="institution">Institut national de recherche en informatique et en automatique</orgName></affiliation></author></analytic></biblStruct></sourceDesc></fileDesc><profileDesc><textClass><keywords scheme="KwdEn" xml:lang="en"><term>Bias</term><term>Gaussian processes</term><term>Markov process</term><term>Maximum likelihood</term><term>Signal processing</term><term>Signal-to-noise ratio</term><term>Speech recognition</term></keywords><keywords scheme="Pascal" xml:lang="fr"><term>Traitement signal</term><term>Rapport signal bruit</term><term>Processus Markov</term><term>Processus gaussien</term><term>Reconnaissance parole</term><term>Maximum vraisemblance</term><term>Erreur systématique</term><term>4360</term><term>4372</term></keywords></textClass></profileDesc></teiHeader><front><div type="abstract" xml:lang="en">In the context of continuous density hidden Markov model (CDHMM) we present a unified maximum likelihood (ML) approach to acoustic mismatch compensation. This is achieved by introducing additive Gaussian biases at the state level in both the mel cepstral and linear spectral domains. Flexible modelling of different mismatch effects can be obtained through appropriate bias tying. A Maximum likelihood approach for joint estimation of both mel cepstral and linear spectral biases from the observed mismatched speech given only one set of clean speech models is presented, where the obtained bias estimates are used for the compensation of clean speech models during decoding. The proposed approach is applied to the recognition of noisy Lombard speech, and significant improvement in the word recognition rate is achieved.</div></front></TEI><affiliations><list><country><li>France</li><li>États-Unis</li></country><region><li>Grand Est</li><li>Lorraine (région)</li></region><settlement><li>Nancy</li></settlement><orgName><li>Centre national de la recherche scientifique</li><li>Institut national de recherche en informatique et en automatique</li><li>Laboratoire lorrain de recherche en informatique et ses applications</li><li>Université de Lorraine</li></orgName></list><tree><country name="France"><region name="Grand Est"><name sortKey="Afify, M" sort="Afify, M" uniqKey="Afify M" first="M." last="Afify">M. Afify</name></region><name sortKey="Haton, J P" sort="Haton, J P" uniqKey="Haton J" first="J.-P." last="Haton">Jean-Paul Haton</name></country><country name="États-Unis"><noRegion><name sortKey="Gong, Y" sort="Gong, Y" uniqKey="Gong Y" first="Y." last="Gong">Y. Gong</name></noRegion></country></tree></affiliations></record>Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Wicri/Lorraine/explor/InforLorV4/Data/Main/Merge
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 00C542 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Main/Merge/biblio.hfd -nk 00C542 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien
|wiki= Wicri/Lorraine
|area= InforLorV4
|flux= Main
|étape= Merge
|type= RBID
|clé= Pascal:98-0082849
|texte= A unified maximum likelihood approach to acoustic mismatch compensation : Application to noisy lombard speech recognition
}}
| This area was generated with Dilib version V0.6.33. |  |