Processing

Please wait...

Settings

Settings

Goto Application

1. WO2022010471 - IDENTIFICATION AND UTILIZATION OF MISRECOGNITIONS IN AUTOMATIC SPEECH RECOGNITION

Publication Number WO/2022/010471
Publication Date 13.01.2022
International Application No. PCT/US2020/041223
International Filing Date 08.07.2020
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 15/01 2013.1
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
15Speech recognition
01Assessment or evaluation of speech recognition systems
G10L 15/065 2013.1
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
15Speech recognition
06Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
065Adaptation
G10L 15/07 2013.1
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
15Speech recognition
06Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
065Adaptation
07to the speaker
Applicants
  • GOOGLE LLC [US]/[US]
Inventors
  • WEISZ, Ágoston
  • MORENO, Ignacio Lopez
  • DOVLECEL, Alexandru
Agents
  • HIGDON, Scott
  • MIDDLETON REUTLINGER
  • SALAZAR, John
  • SHUMAKER, Brantley
  • THRELKELD, Elizabeth
Priority Data
Publication Language English (en)
Filing Language English (EN)
Designated States
Title
(EN) IDENTIFICATION AND UTILIZATION OF MISRECOGNITIONS IN AUTOMATIC SPEECH RECOGNITION
(FR) IDENTIFICATION ET UTILISATION DE RECONNAISSANCES ERRONÉES DANS LA RECONNAISSANCE AUTOMATIQUE DE LA PAROLE
Abstract
(EN) Techniques are disclosed that enable determining and/or utilizing a misrecognition of a spoken utterance, where the misrecognition is generated using an automatic speech recognition (ASR) model. Various implementations include determining a misrecognition based on the spoken utterance and a previous utterance spoken prior to the spoken utterance. Additionally or alternatively, implementations include personalizing an ASR engine for a user based on the spoken utterance and the previous utterance spoken prior to the spoken utterance (e.g., based on audio data capturing the previous utterance and a text representation of the spoken utterance).
(FR) L'invention concerne des techniques qui permettent de déterminer et/ou d'utiliser une reconnaissance erronée d'un énoncé parlé, la reconnaissance erronée étant générée à l'aide d'un modèle de reconnaissance automatique de la parole (ASR). Divers modes de réalisation comprennent la détermination d'une reconnaissance erronée sur la base de l'énoncé parlé et d'un énoncé précédent prononcé avant l'énoncé parlé. De plus ou en variante, des modes de réalisation comprennent la personnalisation d'un moteur ASR pour un utilisateur sur la base de l'énoncé parlé et de l'énoncé précédent prononcé avant l'énoncé parlé (par exemple, sur la base de données audio capturant l'énoncé précédent et d'une représentation textuelle de l'énoncé parlé).
Related patent documents
EP2020750038This application is not viewable in PATENTSCOPE because the national phase entry has not been published yet or the national entry is issued from a country that does not share data with WIPO or there is a formatting issue or an unavailability of the application.
Latest bibliographic data on file with the International Bureau