Method and device for enhancing voice signal

Abstract

The embodiment of the invention discloses a method and a device for enhancing a voice signal. The method comprises the steps of: obtaining a noising voice signal, and carrying out perception weighted filtering on the noising voice signal; converting the noising voice signal subjected to the perception weighted filtering into a frequency domain, carrying out spectrum subtraction and phase synthesis on the noising voice signal in the frequency domain, and converting the voice signal subjected to the spectrum subtraction and phase synthesis into a time domain; and carrying out inverse perceptionweighted filtering on the voice signal subjected to the spectrum subtraction and the phase synthesis to obtain an enhanced voice signal. Through using the invention, the noising voice signal is subjected to the perception weighted filtering, the interference of the noising voice signal to the noise is effectively eliminated, the enhanced voice signal is obtained and the human vision is met.

Description

technical field [0001] The invention relates to the technical field of communication, in particular to a method and device for enhancing voice signals. Background technique [0002] With the development of 3G (3rd Generation, third generation mobile communication system), videophone services have been widely used. While realizing basic communication, the videophone service allows both parties in a call to observe the call scene where the other party is in, which enhances user experience. During a videophone call, in order for the camera to capture the real-time scene image of the caller, both callers need to keep a certain distance from the phone microphone during the call. Therefore, a large number of voice signals collected by the phone microphone are mixed Noise, the introduction of noise reduces the signal-to-noise ratio of the call signal and affects the call quality of the videophone. [0003] In the prior art, in order to reduce the interference of noise to the call...

AI Technical Summary

Problems solved by technology

[0017] However, the process of using the spectral subtraction algorithm to reduce noise in the prior art has the following technical defects: when performing spectral subtraction on noisy speech, the noise spectrum and the speech spectrum cannot be accurately judged. , which also greatly reduces the speech spectrum, which affects the ability of human hearing to perceive the normal speech spectrum.

[0021] During the implementation of Scheme 1 and Scheme 2, the applicant found that in Scheme 1, the implementation process is highly complex, and the control of noise reduction is based on probability distribution, which does not combine the characteristics of human hearing. It is not very satisfactory; in scheme 2, although a pair of α and β values ​​with better effects have been obtained through a large number of experiments, but because the application environment is constantly changing, this method may be better in a certain environment. effect, but in a relatively large environment, its noise reduction control still cannot get satisfactory results

Therefore, in the above scheme for optimizing the performance of speech enhancement by adjusting the intensity of noise reduction in spectral subtraction, the main problem is: when performing spectral subtraction on noisy speech, since the noise spectrum can only be obtained by estimation, The estimated noise spectrum is not accurate enough. In the process of spectrum subtraction, it may be due to the poor control of the reduction intensity that the intensity of the speech spectrum is excessively reduced while reducing the noise, and the enhanced speech signal cannot satisfy human hearing. needs

Images