Unlock instant, AI-driven research and patent intelligence for your innovation.

Speech recognition model training method and device, and speech recognition method and device

A speech recognition model and training method technology, applied in speech recognition, speech analysis, instruments, etc., can solve the problems of limited natural language rules and low recognition accuracy

Active Publication Date: 2020-10-23
BEIJING CENTURY TAL EDUCATION TECH CO LTD
View PDF6 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The end-to-end speech recognition model is trained using the paired annotation corpus of speech and text, and has limited natural language rules, so there is a problem of low recognition accuracy

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Speech recognition model training method and device, and speech recognition method and device
  • Speech recognition model training method and device, and speech recognition method and device
  • Speech recognition model training method and device, and speech recognition method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0099] In the following, only certain exemplary embodiments are briefly described. As those skilled in the art can realize, the described embodiments may be modified in various different ways without departing from the spirit or scope of the present application. Therefore, the drawings and description are to be regarded as illustrative in nature and not restrictive.

[0100] figure 1 A schematic diagram showing a training method of a speech recognition model according to an embodiment of the present application. Such as figure 1 As shown, the method can include:

[0101] Step S11, using the first general corpus to train to obtain a language model module;

[0102] Step S12, using the language model module to obtain an initial speech recognition model, the initial speech recognition model including an acoustic model module, a fusion module and a language model module;

[0103] Step S13: Use speech recognition to label the corpus, train an initial speech recognition model, and obtain a...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides a voice recognition model training method and device, and a voice recognition method and device. The voice recognition model training method comprises the steps: training through a first universal corpus to obtain a language model module; utilizing the language model module to obtain an initial voice recognition model; and training the initial voice recognition model by using the language recognition annotation corpus to obtain a converged voice recognition model. According to the embodiment of the invention, the voice recognition model can grasp sufficient language ruleinformation, and the accuracy of speech recognition is improved.

Description

Technical field [0001] This application relates to the field of speech recognition, and in particular to a training method, speech recognition method and device of a speech recognition model. Background technique [0002] With the rapid development of deep learning, the end-to-end (E2E) model based on deep neural networks has received extensive attention and research in the field of automatic speech recognition, and has become a new research hotspot in the field of speech recognition. The end-to-end speech recognition model is trained on the paired annotation corpus of speech and text, and the natural language laws that it masters are limited. Therefore, there is a problem of low recognition accuracy. Summary of the invention [0003] The embodiments of the present application provide a training method, a voice recognition method and a device for a speech recognition model to solve the problems of related technologies. The technical solutions are as follows: [0004] In the first a...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G10L15/06G10L15/183
CPCG10L15/063G10L15/183
Inventor 李成飞杨嵩徐高鹏
Owner BEIJING CENTURY TAL EDUCATION TECH CO LTD