Voice noise reduction method and device and terminal equipment

A technology for speech noise reduction and speech data, applied in speech analysis, speech recognition, instruments, etc., to solve problems such as inability to accurately transmit user information, noise interference, and reduction of speech recognition accuracy.

Pending Publication Date: 2020-07-07
TCL CORPORATION
View PDF7 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In view of this, the embodiment of the present invention provides a speech noise reduction method, device and terminal equipment to solve the problem of noise interfe...

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
  • Voice noise reduction method and device and terminal equipment
  • Voice noise reduction method and device and terminal equipment
  • Voice noise reduction method and device and terminal equipment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0044] Such as figure 1 As shown, this embodiment provides a speech noise reduction method, which can be applied to terminal devices such as mobile phones, PCs, tablet computers, and smart game consoles. The voice noise reduction method provided in this embodiment includes:

[0045] S101. Acquire first voice data and second voice data; wherein, the first voice data includes noise-free voice data; and the second voice data includes noise voice data.

[0046] In a specific application, a preset number of first voice data and second voice data is obtained for training the generative confrontation network; wherein, the preset number can be specifically set according to the actual situation, and it is required to meet the requirements for the generative confrontation network. training needs. The first voice data refers to real noise-free voice data. The second speech data refers to noise speech data (wherein, the noise speech data includes at least one type of sub-noise speech d...

Embodiment 2

[0061] Such as Figure 5 As shown, this embodiment is a further description of the method steps in the first embodiment. In this embodiment, step 103 includes:

[0062] S1031. Train a generator and a discriminator of the generative adversarial network according to the first speech data and the training data, and obtain a trained generative adversarial network model.

[0063] In a specific application, the training of the discriminator and generator of the generative confrontation network is realized according to the training data and the first speech data, so that the generator can generate more realistic noise-free speech data, so that the discriminator cannot judge the noiseless speech data generated by the generator. Authenticity of the noise voice data; discriminator learns the difference between the first voice data and the training data (i.e. noise-containing voice data), to discriminate the authenticity of the noise-free voice data generated by the generator, and retur...

Embodiment 3

[0070] Such as Figure 8 As shown, this embodiment is a further description of the method steps in the first embodiment. In this embodiment, step S104 includes:

[0071] S1041. Train the discriminator by using the speech recognition data to be processed.

[0072] In a specific application, the discriminator is trained by the speech recognition data to be processed, so that the discriminator can judge whether the speech recognition data contains noise.

[0073] S1042. If the discrimination result of the discriminator is that the speech recognition data to be processed contains noise, train the generator by using the speech recognition data to be processed.

[0074] In a specific application, if the discrimination result of the discriminator is that the speech recognition data to be processed contains noise, the generator is trained by the speech recognition data to be processed, so that the generator generates new speech recognition data to be processed (that is, relatively ...

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 method is suitable for the technical field of voice noise reduction, and provides a voice noise reduction method, a voice noise reduction device and terminal equipment. The method comprises the following steps: obtaining first voice data and second voice data, fusing the first voice data and the second voice data, obtaining training data, training a generative adversarial network according tothe first voice data and the training data, obtaining a trained generative adversarial network model, performing noise reduction on voice recognition data to be processed through the trained generative adversarial network model, and obtaining target voice recognition data. According to the invention, the noise interference in the voice can be reduced, the voice recognition accuracy is improved, and the voice data can accurately transmit the information of the user.

Description

technical field [0001] The invention belongs to the technical field of voice noise reduction, and in particular relates to a voice noise reduction method, device and terminal equipment. Background technique [0002] With the development of the Internet and the wide application of smart mobile terminals, communication by voice has become an important means of communication in people's lives. However, when people use voice to communicate, it is inevitable that there will be various interferences. [0003] During the voice recognition process, if there is noise interference in the voice, the accuracy of voice recognition will be reduced, and the user's information cannot be accurately transmitted. Contents of the invention [0004] In view of this, the embodiment of the present invention provides a speech noise reduction method, device and terminal equipment to solve the problem of noise interference in the speech in the speech recognition process in the prior art, which red...

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
IPC IPC(8): G10L21/0208G10L15/20G10L25/27
CPCG10L21/0208G10L15/20G10L25/27
Inventor 陈明
Owner TCL CORPORATION
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap