A speech signal enhancement method and system based on a logarithmic mmse estimator

A voice signal and estimator technology, applied in the field of signal processing, can solve the problems of ineffective real-time update, poor voice call effect, noise suppression, etc., and achieve the effect of ensuring voice call effect, taking into account processing efficiency, and good noise suppression effect.

Active Publication Date: 2022-02-15
CHENGDU AEROSPACE COMM EQUIP CO LTD
View PDF12 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] In order to overcome the deficiencies of the prior art, the present invention provides a speech signal enhancement method and system based on a logarithmic MMSE estimator, which solves the problems existing in the prior art due to the inability to effectively identify wide and constantly changing (non-stationary) noises , Inability to effectively update and estimate noise in real time, resulting in noise suppression, poor voice call effect, etc.

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 speech signal enhancement method and system based on a logarithmic mmse estimator
  • A speech signal enhancement method and system based on a logarithmic mmse estimator
  • A speech signal enhancement method and system based on a logarithmic mmse estimator

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0052] like Figure 1 to Figure 4 As shown, the present invention updates and estimates the constantly changing (non-stationary) noise with a wide frequency band in real time, by estimating the gain function of the noisy signal in the frequency domain, thereby performing real-time gain on the spectrum of the noisy signal and then Get the noise-suppressed signal. The method can effectively suppress noise that is not stable and has a wide frequency range, thereby improving call quality and experience.

[0053] In the present invention, the gain function of the noisy signal is estimated based on a logarithmic MMSE estimator (MMSE, Minimum Mean Square Error), and noise suppression is performed on the noisy signal. At the same time, a two-stage logarithmic MMSE estimator is cascaded, and the noisy signal is gained twice to minimize the frequency domain energy of the noise, and then the extra noise brought by the two-stage estimator is filtered out through filtering.

[0054] Spec...

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 and provides a voice signal enhancement method and system based on a logarithmic MMSE estimator. In the voice signal enhancement method, the logarithmic MMSE estimator is used to process a noisy signal to obtain a noised signal after gain. The invention solves the problems existing in the prior art, such as noise suppression and poor voice call effect due to the inability to effectively identify wide and constantly changing (non-stationary) noises and to effectively update and estimate noises in real time.

Description

technical field [0001] The invention relates to the technical field of signal processing, in particular to a method and system for enhancing a speech signal based on a logarithmic MMSE estimator. Background technique [0002] In voice communication, it is difficult to avoid the appearance of noise, which affects the quality and intelligibility of voice to a great extent, and it is necessary to suppress the noise through technical methods. Some of these noises have relatively small noise energy and can be removed by the energy threshold, and some noises with high energy frequency domain energy are concentrated in lower frequency bands and can be removed by filters. The environment changes and has a wide frequency band. At this time, this constantly changing (non-stationary) noise suppression cannot be handled by traditional methods. Contents of the invention [0003] In order to overcome the deficiencies of the prior art, the present invention provides a speech signal enha...

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 Patents(China)
IPC IPC(8): G10L21/0208G10L21/0216H04M9/08
CPCG10L21/0208G10L21/0216H04M9/082
Inventor 黄放兰燎张倩李忠秦
Owner CHENGDU AEROSPACE COMM EQUIP 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