Title:
METHOD AND DEVICE FOR PROVIDING VOICE RECOGNITION SERVICE
Document Type and Number:
WIPO Patent Application WO/2020/231181
Kind Code:
A1
Abstract:
A method, performed by the electronic device, of providing a voice recognition service includes obtaining a user call keyword for activating the voice recognition service, based on a first user voice input; generating a user-customized voice database (DB) by inputting the obtained user-customized keyword to a text to speech module; and obtaining a user-customized feature by inputting an audio signal of the user-customized voice DB to a pre-trained wake-up recognition module.
Inventors:
KIM JEONGSU (KR)
CHO KEUNSEOK (KR)
CHO KEUNSEOK (KR)
Application Number:
PCT/KR2020/006300
Publication Date:
November 19, 2020
Filing Date:
May 13, 2020
Export Citation:
Assignee:
SAMSUNG ELECTRONICS CO LTD (KR)
International Classes:
G10L17/24; G10L13/08; G10L15/06; G10L15/18; G10L17/04; G10L17/12
Domestic Patent References:
WO2017217978A1 | 2017-12-21 | |||
WO2017217978A1 | 2017-12-21 |
Foreign References:
US20170018270A1 | 2017-01-19 | |||
KR20190045038A | 2019-05-02 | |||
US20180061402A1 | 2018-03-01 | |||
US20140379334A1 | 2014-12-25 |
Other References:
LUISE VALENTIN RYGAARD, USING SYNTHESIZED SPEECH TO IMPROVE SPEECH RECOGNITION FOR LOW-RESOURCE LANGUAGES, 31 December 2015 (2015-12-31)
UENO SEI ET AL.: "2019 IEEE International Conference On Acoustics, Speech And Signal Processing (ICASSP", 12 May 2019, IEEE, article "Multi-speaker Sequence-to-sequence Speech Synthesis for Data Augmentation in Acoustic-to-word Speech Recognition"
See also references of EP 3888084A4
UENO SEI ET AL.: "2019 IEEE International Conference On Acoustics, Speech And Signal Processing (ICASSP", 12 May 2019, IEEE, article "Multi-speaker Sequence-to-sequence Speech Synthesis for Data Augmentation in Acoustic-to-word Speech Recognition"
See also references of EP 3888084A4
Attorney, Agent or Firm:
Y.P.LEE, MOCK & PARTNERS (KR)
Download PDF: