Unlock instant, AI-driven research and patent intelligence for your innovation.

Method and device for eliminating voice background noise

A background noise and speech technology, applied in the field of speech background noise elimination, can solve problems such as low fidelity and inability to guarantee speech continuity, and achieve the effect of speech signal enhancement and environmental noise removal

Active Publication Date: 2011-03-16
BYD SEMICON CO LTD
View PDF0 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is: the speech background noise elimination method existing in the prior art cannot guarantee the continuity of the speech, introduces new mutation noise in the process of processing the noise, and has low fidelity

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
  • Method and device for eliminating voice background noise
  • Method and device for eliminating voice background noise
  • Method and device for eliminating voice background noise

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0033] figure 1 It is an implementation flowchart of the method for eliminating speech background noise provided by the embodiment of the present invention; figure 1 As shown, a method for eliminating speech background noise provided by an embodiment of the present invention includes the following steps:

[0034] A. Perform characteristic value detection on the external audio voltage signal to obtain a voltage characteristic signal reflecting the changing trend of the external audio voltage signal;

[0035] B. Multiply the voltage characteristic signal with the external audio voltage signal to obtain the product signal; after the external audio voltage signal is processed by eigenvalue detection, the voice signal corresponds to a larger eigenvalue signal, and the noise signal corresponds to a larger eigenvalue signal. Small eigenvalue signal, in the process of multiplying the voltage eigenvalue signal and the external audio voltage signal, the result of multiplying the voice ...

Embodiment 2

[0039] The difference between embodiment two and embodiment one is:

[0040] Step A is specifically: the specific method of detecting the characteristic value of the external audio voltage signal may be: performing effective value detection on the external audio voltage signal to obtain the average power signal of the audio voltage signal;

[0041] Step B is specifically: multiplying the average power signal by the external audio voltage signal to obtain a first product signal;

[0042] Step C is specifically: performing proportional amplification processing on the first product signal and outputting it. In this embodiment example, the multiple of proportional amplification depends on the amplification factor of the pre-amplification process and the degree of attenuation of the noise signal, and is generally designed to be less than 1. multiple.

Embodiment 3

[0044] The difference between embodiment three and embodiment one is:

[0045] Step A is specifically: the specific method for detecting the characteristic value of the external audio voltage signal may be: performing peak detection on the external audio voltage signal to obtain a peak change signal reflecting the peak change trend of the signal;

[0046] Step B is specifically: multiplying the peak change signal by the external audio voltage signal to obtain a second product signal;

[0047] Step C is specifically: performing proportional amplification processing on the second product signal and outputting it. In this embodiment example, the multiple of proportional amplification depends on the amplification factor of the pre-amplification process and the degree of attenuation of the noise signal, and is generally designed to be less than 1. multiple.

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 provides a method for eliminating voice background noise, which comprises the following steps of: A, performing characteristic value detection on an external audio frequency voltage signal to acquire a voltage characteristic signal reflecting the variation trend of the external audio frequency voltage signal; B, multiplying the voltage characteristic signal by the external audio voltage signal to obtain a product; and C, scaling up and outputting the product signal. The invention also provides a device for eliminating voice background noise, which comprises a characteristic value detection unit, a multiplication unit and a scaling up unit, wherein the characteristic value detection unit is used for receiving the external audio frequency voltage signal and performing characteristic value detection on the external audio frequency voltage signal to acquire the voltage characteristic signal reflecting the variation trend of the external audio frequency voltage signal; the multiplication unit is used for multiplying the voltage characteristic signal by the external audio voltage signal to acquire a product signal of the voltage characteristic signal and the external audio voltage signal and output the product signal; and the scaling up unit is used for scaling up and outputting the product signal.

Description

technical field [0001] The invention belongs to the field of speech communication, and in particular relates to a method and device for eliminating speech background noise. Background technique [0002] In the field of voice communication, due to the complex environment of the caller, the audio signal collected by the audio equipment includes not only the caller's voice signal, but also noise signals from other sound sources in the surrounding environment, such as other people's voices, The sound from vehicles, etc. These noise signals seriously interfere with the quality of voice communication. Therefore, while maintaining high-efficiency denoising, the technology of using fewer devices to remove mixed noise signals in voice signals has become the focus of many companies. . [0003] The traditional denoising method is usually to set a noise threshold in the system, and the system recognizes speech above the noise threshold according to the effective value of the input sign...

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): G10L21/02
CPCG10L21/0208
Inventor 李海徐坤平杨云冯卫
Owner BYD SEMICON CO LTD