Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A language recognition method and related device

A language recognition and language technology, applied in the computer field, can solve problems that affect the accuracy of language recognition and fail to meet the needs of multilingual recognition, and achieve the effect of improving accuracy

Active Publication Date: 2022-03-01
TENCENT TECH (SHENZHEN) CO LTD
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] However, due to the diversity of languages, a large number of distinguishing features are involved, and the process of artificially setting distinguishing features cannot meet the recognition needs of multiple languages, which affects the accuracy of language recognition

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
  • A language recognition method and related device
  • A language recognition method and related device
  • A language recognition method and related device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0108]The embodiment of the present application provides a language recognition method and a related device, which can be applied to a system or program that includes a language recognition function in a terminal device, by obtaining input voice; and then inputting the input voice into N parallel neural network models feature extraction layer to obtain N speech feature information, wherein different neural network models are obtained from training in different language training sample sets, and the language categories of the language training samples contained in the same language training sample set are the same, and the feature extraction layer includes at least A deep feedforward sequential memory network and at least one self-attention network, the deep feedforward sequential memory network is used to indicate the temporal features in the input speech, the self-attention network is used to indicate the semantic features in the input speech, the temporal features and the sema...

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 application discloses a language recognition method and a related device. By obtaining the input speech; then input the input speech respectively into the feature extraction layers in N parallel neural network models to obtain N speech feature information; and input the N speech feature information into the data reconstruction layer for speech reconstruction to obtain N A reconstruction loss information; and then determine the language category corresponding to the input speech according to the reconstruction loss information. In this way, the language recognition process based on self-supervised learning is realized. Since the models of each language are independent of each other, the semantic features and timing features between different languages ​​can be automatically mined without setting a large number of distinguishing features, thereby improving the accuracy of language recognition .

Description

technical field [0001] The present application relates to the field of computer technology, in particular to a language recognition method and related devices. Background technique [0002] Automatic Speech Recognition (ASR) is an important direction in the field of computer science and artificial intelligence. It studies various theories and methods for computers to understand human speech, covering multiple disciplines such as acoustics, linguistics, and computer science. . Language recognition is one of the key technologies in speech recognition. It is used to recognize the language category in the speech signal, and transforms the speech recognition problem of cross-language into the speech recognition problem of determining the language. Significance. [0003] Generally, the process of language recognition adopts a supervised classification method based on deep neural network, in which the deep neural network model learns the distinguishing features in the fixed langu...

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 Patents(China)
IPC IPC(8): G10L15/00G10L15/18G10L15/30G10L15/34G06N3/04G06N3/08
CPCG10L15/005G06N3/08G10L15/1815G10L15/30G10L15/34G06N3/044
Inventor 苏丹冯树林
Owner TENCENT TECH (SHENZHEN) CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com