Speech enhancement and model training method and device and electronic equipment

A technology of voice enhancement and electronic equipment, which is applied in the field of data processing, can solve problems such as the trouble of deploying models, and achieve the effect of improving the effect of voice enhancement and good voice enhancement effect

Pending Publication Date: 2022-05-31
BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO LTD
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, it is undoubtedly challenging to learn the features of each sub-band by training a model; and training multiple models to learn the features of each sub-band separately, and then integrating these multiple models is an effective method. method; but it is undoubtedly troublesome to deploy these multiple integrated models on the hardware platform

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 enhancement and model training method and device and electronic equipment
  • Speech enhancement and model training method and device and electronic equipment
  • Speech enhancement and model training method and device and electronic equipment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0034] In order to make the above objects, features and advantages of the present invention more comprehensible, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0035] One of the core concepts of the embodiments of the present invention is to train multiple teacher models in advance using different sub-bands of the speech spectrum, and use multiple teacher models to guide the training of a student model, so that the student model can learn the different sub-bands. features; and then use the student model for speech enhancement, which can achieve a better speech enhancement effect; thus, the speech enhancement effect of the model can be improved without increasing the parameter amount and calculation cost of the speech enhancement model.

[0036] The following first describes how to train the student model.

[0037] refer to figure 1 , which shows a flow chart of the steps of an embodimen...

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 embodiment of the invention provides a model training and voice enhancement method and device and electronic equipment, and the method comprises the steps: obtaining voice data; performing voice enhancement on the voice data according to a preset voice enhancement model, the preset voice enhancement model being a student model guided and trained by a teacher model, and the teacher model comprising a plurality of models trained based on different sub-bands of a voice spectrum, further, the student model can learn features on different sub-bands; therefore, a better speech enhancement effect can be achieved, and the speech enhancement effect of the model is improved on the premise that the parameter quantity and the calculation cost of the speech enhancement model are not increased.

Description

technical field [0001] The invention relates to the technical field of data processing, in particular to a speech enhancement and model training method, device and electronic equipment. Background technique [0002] With the rapid development of communication technology, terminals such as mobile phones and tablet computers are becoming more and more popular, which brings great convenience to people's life, study and work. Users usually use the terminal to input voice commands to perform voice photography, voice search, etc.; also use the terminal to play voice data (such as music, video, recording). Among them, in order to enable the terminal to better execute the user's voice commands and play voice data with higher quality, the terminal can use voice enhancement technology to suppress and reduce noise interference in voice data after collecting voice commands or before playing voice data, Extract useful speech data from noisy background. [0003] At present, the method o...

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): G10L21/02G10L21/0232
CPCG10L21/02G10L21/0232
Inventor 文仕学郝翔
Owner BEIJING SOGOU TECHNOLOGY DEVELOPMENT CO LTD
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