Comparative experiments of several adaptation approaches to noisy speech recognition using stochastic trajectory models

Comparative experiments of several adaptation approaches to noisy speech recognition using stochastic trajectory models

Source :

• Speech Communication [ 0167-6393 ] ; 1996.

Descripteurs français

• Pascal (Inist)
  • Etude comparative, Filtrage, Modèle stochastique, Reconnaissance parole, Réduction bruit, Régression linéaire, Traitement parole.

English descriptors

• KwdEn :
  • Comparative study, Filtering, Linear regression, Noise reduction, Speech processing, Speech recognition, Stochastic model.
• Teeft :
  • Acoustic, Acoustic models, Acoustic training, Adaptation, Adaptation class, Adaptation corpus, Adaptation data, Adaptation sentences, Adaptation trajectory, Additive noise, Autobus noises, Auxiliary function, Base element, Base elements, Base transformation, Best results, Cambridge university, Cepstral, Cepstral domain, Cepstrum, Clean speech, Clean speech signal, Clean speech sims, Compemolle, Component trajectory, Confidence intervals, Continuous speech recognition task, Covariance, Covariance matrix, Diagonal covariance matrices, Diagonal matrices, Different approaches, Different symbols, Differentes approches, Discrete cosine, Engineering department, Estimation, Estimation errors, Free parameters, Function communication, Gaussian, Gaussian distribution, Gaussian noise, General tendency, Gong, Grand vector, Haton, High snrs, Ieee internat, Ieee trans, Label observations, Language processing, Large number, Last column, Linear combination, Linear power, Linear regression, Linear regression framework, Linear regression scheme, Linear transformation, Logarithmic domain, Lynx helicopter, Matrix, Medium snrs, Mixture component, Model combination, Model combination framework, Noise adaptation, Noise adaptation schemes, Noise model, Noise observations, Noise types, Noisy, Noisy environment, Noisy environments, Noisy speech, Noisy speech cepstrum vector, Noisy speech recognition, Noisy speech recognition approaches, Noisy trajectory, Normal distribution, Other approaches, Other methods, Parameter, Phone models, Phoneme, Phoneme symbol, Phonetic label, Phonetic level, Precision matrix, Proc, Recognition rate, Recognition results, Reference environment, Regression, Same transformation matrix, Search stage, Similar approach, Siohan, Small adaptation corpus, Snrs, Speaker adaptation, Speech adaptation corpus, Speech communicarion, Speech communication, Speech enhancement, Speech recognition, Speech recognition approaches, Speech recognition system, Speech recognizers, Speech signal process, Square matrix, Stm, Stochastic, Stochastic trajectory models, Technical report, Testing conditions, Theoretical framework, Total number, Training corpus, Training sentences, Trajectory, Transformation matrices, Transformation matrix, Uncompensated system, Utterance segments, Word accuracy, Word error reduction, Words vocabulary.

Abstract

• https://api.istex.fr/ark:/67375/6H6-58RN2NFV-6/fulltext.pdf

Affiliations:

• France
• Grand Est, Lorraine (région)
• Nancy, Vandæuvre-lès-Nancy
• Centre national de la recherche scientifique, Institut national de recherche en informatique et en automatique, Laboratoire lorrain de recherche en informatique et ses applications, Université de Lorraine

Links toward previous steps (curation, corpus...)

• to stream Istex, to step Corpus: 001512
• to stream Istex, to step Curation: 001495
• to stream Istex, to step Checkpoint: 002864
• to stream Main, to step Merge: 00C784
• to stream PascalFrancis, to step Corpus: 000D35
• to stream PascalFrancis, to step Curation: 000B56
• to stream PascalFrancis, to step Checkpoint: 000C98
• to stream Main, to step Merge: 00C985
• to stream Main, to step Curation: 00BF63

Le document en format XML

Pour manipuler ce document sous Unix (Dilib)

EXPLOR_STEP=$WICRI_ROOT/Wicri/Lorraine/explor/InforLorV4/Data/Main/Exploration

HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 00BF63 | SxmlIndent | more

Ou

HfdSelect -h $EXPLOR_AREA/Data/Main/Exploration/biblio.hfd -nk 00BF63 | SxmlIndent | more

Pour mettre un lien sur cette page dans le réseau Wicri

{{Explor lien
   |wiki=    Wicri/Lorraine
   |area=    InforLorV4
   |flux=    Main
   |étape=   Exploration
   |type=    RBID
   |clé=     ISTEX:5AE9758FB450A76F8EC2B0DEA437CEC603F1A1BB
   |texte=   Comparative experiments of several adaptation approaches to noisy speech recognition using stochastic trajectory models
}}

This area was generated with Dilib version V0.6.33. Data generation: Mon Jun 10 21:56:28 2019. Site generation: Fri Feb 25 15:29:27 2022