Using confidence measure for keyword detection in continuous speech recognition - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Communication Dans Un Congrès Année : 2004

Using confidence measure for keyword detection in continuous speech recognition

Dominique Fohr
Gérard Chollet

Résumé

This paper deals with the problem of detection keywords/rejection out-of-vocabulary in continuous speech recognition. Two different techniques based on confidence measures are investigated to improve the detection of keywords using a grammar founded on loop of phones. Confidence measures are computed from phone level information provided by a Hidden Markov model based speech recognizer. We use two kinds of likelihood as ratio and distance, and theirs normalised forms to compute a confidence measures for each word. All confidence measures are are evaluated using the French SPEECHDAT database. The Figure-Of-Merite (FOM) for the normalized likelihood ratio is about $68.2\%$ compared to $71.5\%$ obtained by the normalized likelihood distance
Fichier non déposé

Dates et versions

inria-00100205 , version 1 (26-09-2006)

Identifiants

  • HAL Id : inria-00100205 , version 1

Citer

Yassine Benayed, Dominique Fohr, Jean-Paul Haton, Gérard Chollet. Using confidence measure for keyword detection in continuous speech recognition. Conférence Internationale sur l'accès Intelligent aux Documents Multimédia sur l'Internet - Medinet'04, 2004, Tozeur, Tunisie, 10 p. ⟨inria-00100205⟩
154 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More