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

A multichannel speech enhancement method using postfilter

A post-filter and speech enhancement technology, applied in speech analysis, instruments, etc., can solve problems such as inability to be practical, irrelevant and irrelevant signals and background noise, and achieve the effect of improving the effect.

Inactive Publication Date: 2004-08-25
INST OF ACOUSTICS CHINESE ACAD OF SCI
View PDF0 Cites 23 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0015] 1. Signal and background noise are uncorrelated
[0016] 2. The noise recorded by each channel is also irrelevant
[0023] However, this method is based on the assumption that the noises recorded by each channel are also uncorrelated, but the cross-correlation of the noise of each channel can be basically ignored only in the case of high frequency, and in the case of low frequency, each The cross-correlation of channel noise is more obvious and cannot be ignored, so this method cannot be practical

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 multichannel speech enhancement method using postfilter
  • A multichannel speech enhancement method using postfilter
  • A multichannel speech enhancement method using postfilter

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0041] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0042] For the signal model x given by formulas (1) and (2) i (t) and x j (t) remove the time delay τ i , τ j After doing the Fourier transform, we get

[0043] X ^ i ( f ) = S ( f ) + N i ( f ) e j 2 π W f τ i - - - - - ( 6 )

[0044] ...

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 kind of voice enhancing method which uses postposition filter, it includes following steps: 1) calculates the time delay of voice signal in each channel; 2) aligns each channel in time domain through delay compensation; 3) carries on wave-packet formation to signal in each channel with wave-packet producer; 4) estimates the self power spectrum and noise including signal self power spectrum of purified voice signal, gets the frequency response function of Wiener filter; it obtains the self power spectrum estimation through eliminating the noise self power spectrum estimation from noise including signal cross-power spectrum; 5) uses the postposition Wiener filter to carry on filter to wave-packet producer, realizes the enhancement of voice. The invention considers the correlation of each channel noise, this accords the reality condition, especially it can eliminate the noise in low frequency , enhances the effect of voice enhancing.

Description

technical field [0001] The present invention relates to the field of computer speech signal processing, more specifically, the present invention relates to a multi-channel speech enhancement method using a post filter Background technique [0002] Speech enhancement is a selective signal processing technology, which mainly solves the problem of extracting the target speech signal as pure as possible from the speech signals polluted in different ways. One purpose of speech enhancement is to improve the sense of hearing and intelligibility of speech signals, and it can be used in communication, hearing aid, listening, audio-visual conferencing and other fields. In addition, with the development of speech recognition technology, a high recognition rate can be achieved in a quiet environment, but the degradation of the recognition rate is more serious in a noisy environment. Therefore, speech enhancement, as a means of speech recognition front-end processing, is a very active a...

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): G10L19/26G10L21/0208
Inventor 杜利民阎兆立
Owner INST OF ACOUSTICS CHINESE ACAD OF SCI
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