Noise suppression method and noise suppressor

A noise suppression and noise technology, applied in the field of speech processing, can solve the problem of long noise suppression convergence time, and achieve the effect of improving timeliness and accuracy

Active Publication Date: 2014-04-16
MORNINGCORE TECH CO LTD
View PDF3 Cites 15 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0017] The technical problem to be solved by the present invention is to provide a noise suppression method and a noise su

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
  • Noise suppression method and noise suppressor
  • Noise suppression method and noise suppressor
  • Noise suppression method and noise suppressor

Examples

Experimental program
Comparison scheme
Effect test

Example Embodiment

[0034] In summary, the embodiment of the present invention proposes a noise power spectrum backward prediction technology, through which the noise power spectrum can be pre-estimated before the noise processing, and the noise power spectrum can be estimated during the processing by using the pre-estimation result. Tracking is performed to estimate the convergent noise power spectral density before processing noise suppression. In this way, the noise suppressor does not substantially require convergence time.

[0035] figure 2 A structural diagram of a noise suppressor according to an embodiment of the present invention is shown. Reference figure 2 As shown, the noise suppressor includes first and second short-time Fourier analysis units 201, 202, noise estimation unit 203, noise tracking unit 204, signal-to-noise ratio estimation unit 205, gain calculation unit 206, noise suppression unit 207, and Short-time Fourier synthesis unit 208.

[0036] image 3 Shows a flowchart of a n...

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 relates to a noise suppression method and a noise suppressor. The noise suppression method comprises the steps that short-time Fourier analysis is conducted on at least part of sound source data, so that a first short-time frequency spectrum is obtained; pre-estimation is conducted on a noise power spectrum by means of the first short-time frequency spectrum, so that a pre-estimation noise power spectrum is obtained; short-time Fourier analysis is conducted on the whole sound source data, so that a second short-time frequency spectrum is obtained; the pre-estimation noise power spectrum is used as an initial power spectrum density, and noise tracking is conducted on the second short-time frequency spectrum, so that noise estimation energy is obtained; the signal to noise ratio is estimated by means of the noise estimation energy and the second short-time frequency spectrum; the attenuation gain of noise suppression is calculated according to the signal to noise ratio; the noise suppression is conducted according to the attenuation gain.

Description

technical field [0001] The invention relates to a voice processing, in particular to a noise suppression method and a noise suppressor. Background technique [0002] In the environment with noisy background, the recognition rate of speaker recognition will be poor due to noise interference. Therefore, there will be a noise suppression processing process at the front end of the speaker recognition system. In traditional noise suppression processing, the process is as follows figure 1 As shown, the process is as follows: [0003] In step 101, windowing and short-time Fourier transform are first performed on the noisy speech to obtain the short-time Fourier transform spectrum: [0004] X STFT [ k , m ] = Σ n = - ∞ n = ∞ ...

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): G10L21/0232
Inventor 许云峰王彦谢单辉
Owner MORNINGCORE TECH 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