Title:
VOICE RECOGNITION METHOD AND APPARATUS, DEVICE AND COMPUTER READABLE STORAGE MEDIUM
Document Type and Number:
WIPO Patent Application WO/2024/022541
Kind Code:
A1
Abstract:
Embodiments of the application provide a voice recognition method and apparatus, a device and a computer readable storage medium. The method comprises: obtaining audio data to be recognized; according to the audio data and a trained target acoustic neural network, determining a phoneme sequence corresponding to the audio data; the target acoustic neural network being obtained by training by means of unsupervised sample data and supervised sample data which carries a text label; inputting the phoneme sequence into a trained language model for processing, and outputting a voice recognition result of the audio data. In the embodiments of the application, the target acoustic neural network is trained by utilizing a large amount of unsupervised sample data and supervised sample data which carries a text label, so as to effectively solve the problem of insufficient estimation of the voice feature space caused by insufficient amount of data of the supervised sample, improve the recognition precision of the trained target acoustic neural network, and improve the subsequent voice recognition effect.
Inventors:
QI XIN (CN)
Application Number:
PCT/CN2023/121092
Publication Date:
February 01, 2024
Filing Date:
September 25, 2023
Export Citation:
Assignee:
SF TECH CO LTD (CN)
International Classes:
G10L15/06; G10L15/02; G10L15/16
Foreign References:
| CN114783464A | 2022-07-22 | |||
| CN114399995A | 2022-04-26 | |||
| CN114596844A | 2022-06-07 | |||
| CN107240395A | 2017-10-10 | |||
| US20180075844A1 | 2018-03-15 |
Attorney, Agent or Firm:
BEIJING BRIGHT IP AGENCY CO., LTD. (CN)
Download PDF: