Title:
SPEECH RECOGNITION METHOD AND APPARATUS, DEVICE, AND STORAGE MEDIUM
Document Type and Number:
WIPO Patent Application WO/2022/078146
Kind Code:
A1
Abstract:
A speech recognition method and apparatus, a device, and a storage medium, relating to the field of speech recognition. Speech recognition is performed using attention coding and time series decoding methods, and attention coding is performed on a speech feature matrix, such that the parallel computing of a GPU can be efficiently utilized. Decoding coded vectors according to the positions of the coded vectors in a coded matrix can reduce the amount of parameters, and the use of memory features of a speech recognition model to perform prediction according to time-series related coded vectors can improve speech recognition accuracy.
Inventors:
ZHANG XILIN (CN)
LIU BO (CN)
WANG HAIPENG (CN)
MA JIANXIONG (CN)
ZHENG PING (CN)
LIU BO (CN)
WANG HAIPENG (CN)
MA JIANXIONG (CN)
ZHENG PING (CN)
Application Number:
PCT/CN2021/118514
Publication Date:
April 21, 2022
Filing Date:
September 15, 2021
Export Citation:
Assignee:
TENCENT TECH SHENZHEN CO LTD (CN)
International Classes:
G10L15/06; G10L15/02
Domestic Patent References:
| WO2020101743A1 | 2020-05-22 |
Foreign References:
| CN111933115A | 2020-11-13 | |||
| CN110473529A | 2019-11-19 | |||
| CN111524503A | 2020-08-11 | |||
| CN108630199A | 2018-10-09 | |||
| CN111145728A | 2020-05-12 | |||
| CN110444203A | 2019-11-12 |
Attorney, Agent or Firm:
BEIJING SAN GAO YONG XIN INTELLECTUAL PROPERTY AGENCY CO., LTD. (CN)
Download PDF: