Voice signal enhancement method

A voice signal and signal technology, applied in voice analysis, instruments, electrical components, etc., can solve the problems of poor judgment accuracy, difficult to eliminate music noise, affecting the accuracy of voice noise estimation, etc., to achieve accurate judgment, improve accuracy, improve The effect of the voice enhancement effect

Pending Publication Date: 2021-04-13
江苏集萃智能集成电路设计技术研究所有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Aiming at the problem that the existing voice noise elimination method in the prior art has poor judgment accuracy on the existence of voice, affects the accuracy of voice noise estimation, and produces music noise that is difficult to eliminate, the present invention provides a voice signal enhancement method , which can improve the accuracy of speech existence judgment, thereby improving the accuracy of estimating noise, and can effectively eliminate musical noise in speech signals

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 signal enhancement method
  • Voice signal enhancement method
  • Voice signal enhancement method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0053] See figure 1 , a speech signal enhancement method, the speech enhancement implementation steps are as follows:

[0054] Step a: collect the noisy speech signal y(n), the noisy speech signal y(n) is a time domain signal, the noise contained in the noisy speech signal y(n) is additive noise d(n) and the pure speech is x( n), wherein, n is a discrete time;

[0055] Step b: converting the time-domain signal into a frequency-domain signal by using a weighted lap-and-add (WOLA) analysis filter bank, the frequency-domain signal Y(k)=D(k)+X(k);

[0056] Step c: Update the estimated value of the additive noise power spectrum using the noise estimation algorithm

[0057] Step d: Use the multi-subband spectral subtraction algorithm to perform spectral subtraction of the power spectrum of the noisy speech to obtain the estimated value of the power spectrum of the pure speech

[0058] Firstly, the amplitude spectrum Y of the noisy speech of each subband i (k) do smoothing: ...

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 discloses a voice signal enhancement method, which can improve the accuracy of voice existence judgment so as to improve the accuracy of noise estimation and effectively eliminate music noise in a voice signal. The method comprises the following steps of: acquiring a noisy voice signal which comprises additive noise, converting a time domain signal into a frequency domain signal by adopting a weighted splicing addition filter bank, updating a power spectrum estimation value of the additive noise signal by adopting a noise estimation algorithm provided by the invention, and performing spectral subtraction of a power spectrum on noisy voice by adopting a multi-sub-band spectral subtraction algorithm to obtain a power spectrum estimation value of pure voice, and restoring the frequency domain signal of the pure voice by using the phase of the noisy speech and the estimated value of the amplitude spectrum of the pure voice, and converting the frequency domain signal into a time domain signal again by using a weighted lap joint addition filter bank to obtain the pure voice after voice enhancement.

Description

technical field [0001] The invention relates to the technical field of voice signal processing, in particular to a voice signal enhancement method. Background technique [0002] In speech systems, noise is often the most annoying, so speech enhancement is one of the most critical technologies in speech processing systems. Speech enhancement consists of two parts: noise estimation and noise removal. [0003] Traditional noise estimation is noise estimation based on the existence of speech. This method only performs noise estimation in the non-speech segment, and it only has a good effect on stationary noise, and cannot accurately track noise for non-stationary noise. Based on this, S.Rangachari proposed a noise estimation method with a small amount of calculation and is more suitable for hardware implementation, the minimum-controlled recursive averaging algorithm (Minima-Controlled Recursive Averaging, MCRA-2), which first uses the minimum tracking method Track the noise i...

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): G10L21/0216G10L25/21H04R25/00
CPCG10L21/0216G10L25/21H04R25/50
Inventor 王昂肖宛昂兰玉燕周维新马万康
Owner 江苏集萃智能集成电路设计技术研究所有限公司
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