Processing

Please wait...

Settings

Settings

Goto Application

1. WO2007088299 - CONFIDENCE MEASURE ASSOCIATED WITH AN AUTOMATIC VOICE RECOGNITION HYPOTHESIS

Publication Number WO/2007/088299
Publication Date 09.08.2007
International Application No. PCT/FR2007/050698
International Filing Date 26.01.2007
IPC
G10L 15/08 2006.01
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
15Speech recognition
08Speech classification or search
CPC
G10L 15/08
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
15Speech recognition
08Speech classification or search
Applicants
  • FRANCE TELECOM [FR]/[FR] (AllExceptUS)
  • CHARLET, Delphine [FR]/[FR] (UsOnly)
  • JOUVET, Denis [FR]/[FR] (UsOnly)
Inventors
  • CHARLET, Delphine
  • JOUVET, Denis
Agents
  • LAPOUX, Roland
Priority Data
060088831.01.2006FR
Publication Language French (FR)
Filing Language French (FR)
Designated States
Title
(EN) CONFIDENCE MEASURE ASSOCIATED WITH AN AUTOMATIC VOICE RECOGNITION HYPOTHESIS
(FR) MESURE DE CONFIANCE ASSOCIEE A UNE HYPOTHESE DE RECONNAISSANCE VOCALE AUTOMATIQUE
Abstract
(EN)
The invention relates to a hypothesis generation module (MGH) which generates an acceptance hypothesis and a competition hypothesis relating to a voice segment delivered by a speaker during automatic voice recognition. In order to associate a confidence measure with the acceptance hypothesis generated, an evaluation module (MEV) determines similarity between hypotheses, dependent on a distance between the acceptance and competition hypotheses, and evaluates a confidence measure which is associated with the acceptance hypothesis as a function of likelihood between the voice segment and the acceptance and competition hypotheses and the determined similarity between hypotheses.
(FR)
Un module de génération d'hypothèses (MGH) génère une hypothèse d'acceptation et une hypothèse de concurrence relatives à un segment vocal prononcé par un locuteur lors d'une reconnaissance vocale automatique. Pour associer une mesure de confiance à l'hypothèse d'acceptation générée, un module d'évaluation (MEV) détermine une similarité d'hypothèses dépendant d'une distance entre les hypothèses d'acceptation et de concurrence et évalue une mesure de confiance associée à l'hypothèse d'acceptation en fonction de vraisemblances entre le segment vocal et les hypothèses d'acceptation et de concurrence, et de la similarité d'hypothèses déterminée.
Latest bibliographic data on file with the International Bureau