Processing

Please wait...

Settings

Settings

Goto Application

1. WO2017202016 - VOICE WAKE-UP METHOD AND DEVICE

Publication Number WO/2017/202016
Publication Date 30.11.2017
International Application No. PCT/CN2016/111367
International Filing Date 21.12.2016
IPC
G10L 15/22 2006.01
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
CPC
G10L 15/04
GPHYSICS
10MUSICAL INSTRUMENTS; ACOUSTICS
LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
15Speech recognition
04Segmentation; Word boundary detection
G10L 15/063
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
063Training
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
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 2015/088
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
088Word spotting
G10L 2015/223
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
223Execution procedure of a spoken command
Applicants
  • 百度在线网络技术(北京)有限公司 BAIDU ONLINE NETWORK TECHNOLOGY (BEIJING) CO., LTD. [CN]/[CN]
Inventors
  • 袁斌 YUAN, Bin
Agents
  • 北京清亦华知识产权代理事务所(普通合伙) TSINGYIHUA INTELLECTUAL PROPERTY LLC
Priority Data
201610357702.426.05.2016CN
Publication Language Chinese (ZH)
Filing Language Chinese (ZH)
Designated States
Title
(EN) VOICE WAKE-UP METHOD AND DEVICE
(FR) PROCÉDÉ ET DISPOSITIF DE RÉVEIL VOCAL
(ZH) 语音唤醒方法和装置
Abstract
(EN)
The invention discloses a voice wake-up method and a device. The voice wake-up method comprises steps as follows: acquiring a to-be-processed voice signal (S11); decoding the voice signal according to a pre-generated searching space to obtain a voice recognition result, wherein the searching space comprises a path where an inverse model is located, and the inverse model comprises a first inverse model generated by training word segmentation results of wake-up words (S12); determining whether the words with the preset number of characters contains at least part of characters in the wake-up words when the preset number of characters before the voice recognition result is obtained (S13); and directly determining cancellation of wake-up operation if at least part of characters in the wake-up words are not contained, and ending the decoding of the voice signal (S14). The method has the advantages that the mistaken wake-up rate can be reduced, and the power consumption can be reduced.
(FR)
La présente invention concerne un procédé et un dispositif de réveil vocal. Le procédé de réveil vocal comprend les étapes suivantes : acquérir un signal vocal à traiter (S11) ; décoder le signal vocal selon un espace de recherche pré-généré afin d'obtenir un résultat de reconnaissance vocale, l'espace de recherche comprenant un chemin où se trouve un modèle inversé, et le modèle inversé comprenant un premier modèle inversé généré par des résultats de segmentation de mots d'apprentissage de mots de réveil (S12) ; déterminer si les mots avec le nombre prédéterminé de caractères contiennent au moins une partie de caractères dans les mots de réveil lorsque le nombre prédéterminé de caractères avant le résultat de reconnaissance vocale est obtenu (S13) ; et déterminer directement l'annulation de l'opération de réveil si au moins une partie des caractères dans les mots de réveil n'est pas contenue, et terminer le décodage du signal vocal (S14). Le procédé présente les avantages suivants : le taux de réveil erroné peut être réduit, et la consommation d'énergie peut être réduite.
(ZH)
一种语音唤醒方法和装置,该语音唤醒方法包括:获取待处理的语音信号(S11);根据预先生成的搜索空间,对所述语音信号进行解码,得到语音识别结果,其中,所述搜索空间包括反模型所在路径,所述反模型包括第一反模型,所述第一反模型根据对唤醒词的分词结果训练生成(S12);当获取到所述语音识别结果的前面的预设个数的字时,判断所述预设个数的字中是否包含唤醒词中的至少部分字(S13);如果不包含,则直接确定不唤醒,结束对所述语音信号的解码(S14)。该方法能够降低误唤醒率和降低功耗。
Also published as
Latest bibliographic data on file with the International Bureau