Dual-channel voice enhancement method based on noise power spectral density

A technology of noise power spectrum and power spectral density, which is applied in speech analysis, instruments, etc., can solve the problems of irrelevant noise signals and inability to achieve good results, and achieve distortion-free, good application prospects and strong portability Effect

Inactive Publication Date: 2018-02-09
GUILIN UNIV OF ELECTRONIC TECH
View PDF5 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] In existing small mobile devices, a single microphone or a dual-microphone array is often used to deal with coherent and incoherent noise in the surrounding environment, while traditional methods including spectral subtraction, Wiener filtering and methods based on statistical models cannot achieve great results. Good results
T...

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
  • Dual-channel voice enhancement method based on noise power spectral density
  • Dual-channel voice enhancement method based on noise power spectral density
  • Dual-channel voice enhancement method based on noise power spectral density

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0051] refer to Figure 7 , a two-channel speech enhancement method based on noise power spectral density, comprising the steps of:

[0052] 1) Short-time Fourier transform is performed on the speech signals received by the two channels: Assume that the speech signals received by the main microphone and the secondary microphone are as follows:

[0053] x 1 (t)=h 1 the s 1 (t)+n 1 (t)

[0054] x 2 (t)=h 2 the s 2 (t-τ)+n 2 (t)

[0055] where x 1 (t) and x 2 (t) is the noisy speech signal received by the main microphone and the secondary microphone at time t, image 3 and Figure 4 The time-domain waveform diagram of the noisy speech signal of the main microphone and the secondary microphone respectively; s1 (t) and s 2 (t) is the source signal at time t, figure 1 and figure 2 The time-domain waveform diagram of the pure voice signal of the main microphone and the secondary microphone respectively; h 1 and h 2 Transmission attenuation of primary and secondary ...

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 dual-channel voice enhancement method based on noise power spectral density, and the method is characterized in that the method comprises the following steps: 1), carrying out the short-time Fourier transform of voice signals received by two channels; 2), constructing a cross-power spectrum reduction filter; 3), constructing a spectrum correction filter, and carrying outthe preliminary estimation of the power spectral density of ta noise signal; 4), carrying out the accurate estimation of noise in the noise-containing voice signals. The method can reduce the coherentnoise residual, can improve the inhibition of non-coherent noise, is high in transportability, and is very good in application prospect for small-size intelligent mobile equipment.

Description

technical field [0001] The invention relates to the technical field of speech enhancement, in particular to a dual-channel speech enhancement method based on noise power spectral density. Background technique [0002] In existing small mobile devices, a single microphone or a dual-microphone array is often used to deal with coherent and incoherent noise in the surrounding environment, while traditional methods including spectral subtraction, Wiener filtering and methods based on statistical models cannot achieve great results. Good results. The classic dual-microphone array denoising technology based on frequency-domain coherence function calculates the cross-power spectral density of the signal, and constructs a coherent filter from the power spectral density for filtering, but this method is limited in that the noise signal between the two microphones is required is irrelevant. Contents of the invention [0003] The purpose of the present invention is to provide a two-...

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): G10L21/0208G10L21/0216G10L21/0264G10L25/27
CPCG10L21/0208G10L21/0216G10L21/0264G10L25/27G10L2021/02165
Inventor 曾庆宁毛维
Owner GUILIN UNIV OF ELECTRONIC TECH
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