Processing

Please wait...

Settings

Settings

Goto Application

1. WO2022119705 - DECAYING AUTOMATED SPEECH RECOGNITION PROCESSING RESULTS

Publication Number WO/2022/119705
Publication Date 09.06.2022
International Application No. PCT/US2021/059588
International Filing Date 16.11.2021
IPC
G10L 15/22 2006.1
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
15Speech recognition
22Procedures used during a speech recognition process, e.g. man-machine dialog
G10L 15/30 2013.1
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
15Speech recognition
28Constructional details of speech recognition systems
30Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
G10L 25/78 2013.1
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
25Speech or voice analysis techniques not restricted to a single one of groups G10L15/-G10L21/129
78Detection of presence or absence of voice signals
CPC
G10L 15/22
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
15Speech recognition
22Procedures used during a speech recognition process, e.g. man-machine dialogue
G10L 15/30
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
15Speech recognition
28Constructional details of speech recognition systems
30Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
G10L 2015/228
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
15Speech recognition
22Procedures used during a speech recognition process, e.g. man-machine dialogue
226using non-speech characteristics
228of application context
G10L 25/78
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
25Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
78Detection of presence or absence of voice signals
Applicants
  • GOOGLE LLC [US]/[US]
Inventors
  • SHARIFI, Matthew
  • CARBUNE, Victor
Agents
  • KRUEGER, Brett, A.
Priority Data
17/111,46703.12.2020US
Publication Language English (en)
Filing Language English (EN)
Designated States
Title
(EN) DECAYING AUTOMATED SPEECH RECOGNITION PROCESSING RESULTS
(FR) DÉCOMPOSITION DE RÉSULTATS DE TRAITEMENT DE RECONNAISSANCE AUTOMATIQUE DE LA PAROLE
Abstract
(EN) A method (300) for decaying speech processing includes receiving, at a voice- enabled device (110), an indication of a microphone trigger event (202) indicating a possible interaction with the device through speech where the device has a microphone (116) that, when open, is configured to capture speech. In response to receiving the indication of the microphone trigger event, the method also includes instructing the microphone to open or remain open for a duration window (212) to capture an audio stream (16) and providing the audio stream captured by the open microphone to a speech recognition system (150). During the duration window, the method further includes decaying a level (222) of the speech recognition processing based on a function of the duration window and instructing the speech recognition system to use the decayed level (204, 222) of speech recognition processing over the audio stream.
(FR) L’invention concerne un procédé (300) destiné à décomposer le traitement de la parole qui consiste à recevoir, au niveau d’un dispositif activé par la parole (110), une indication d’un événement de déclenchement (202) de microphone indiquant une interaction possible avec le dispositif par la parole où le dispositif a un microphone (116) qui, lorsqu’il est ouvert, sert à capturer la parole. En réponse à la réception de l’indication de l’événement de déclenchement de microphone, le procédé consiste également à ordonner au microphone de s’ouvrir ou de rester ouvert pendant une fenêtre de durée (212) pour capturer un flux audio (16) et à transmettre le flux audio capturé par le microphone ouvert à un système de reconnaissance de la parole (150). Durant la fenêtre de durée, le procédé consiste en outre à décomposer un niveau (222) du traitement de reconnaissance de la parole sur la base d’une fonction de la fenêtre de durée et à ordonner au système de reconnaissance de la parole d’utiliser le niveau décomposé (204, 222) de traitement de reconnaissance de la parole sur le système audio.
Related patent documents
Latest bibliographic data on file with the International Bureau