Speech Enhancement and Frequency Response Compensation Fusion Method in Digital Hearing Aid

Abstract

The invention provides a speech enhancing and frequency response compensation fusion method in a digital hearing-aid. The speech enhancing and frequency response compensation fusion method includes the steps of (1) obtaining estimated noise and initial enhanced speech with an MCRA method, respectively carrying out filtering processing on the estimated noise and the initial enhanced speech through a gammatone filter, dividing a signal into M frequency bands through the perception mechanism of the cochleas to the signal, and meanwhile obtaining a time frequency expression mode of the signal, (2) computing masking threshold values of the frequency bands through factors such as the audio masking characteristic of the human ears and the frequency band signal to noise ratios, (3) dynamically computing masking values of noise-contained speech in a time frequency domain through a hearing curve of a hearing disorder patient, and processing speech enhancing and frequency response compensation at the same time, and (4) synthesizing output speech of the heating-aid through the masking values. According to the speech enhancing and frequency response compensation fusion method, the working mechanism of the human ears is sufficiently used, the speech characteristics are kept, music noise introduced through a spectral subtraction method is eliminated, the speech intelligibility of output signals of the hearing-aid is greatly improved, and the low complexity and the low power consumption are achieved.

Description

technical field [0001] The invention belongs to the technical field of speech signal processing and relates to two key speech signal processing technologies of speech enhancement and frequency response compensation in digital hearing aids. Background technique [0002] Deafness has become a worldwide problem, but under the current medical conditions, for most deaf people, choosing a suitable digital hearing aid is the best way to effectively improve their hearing level. With the rapid development of science and technology in recent decades, the performance of digital hearing aids has been continuously improved. The basic working principles of digital hearing aids, such as figure 2 As shown, the external sound signal enters the microphone to convert sound energy into electrical energy, and then converts it into a digital signal through an analog / digital converter, and then uses a preset algorithm in the digital microprocessor and digital filter to process this digital signal....

AI Technical Summary

Problems solved by technology

[0007] The present invention aims at the existing digital hearing aid speech enhancement algorithm in a non-stationary noise environment, a large amount of background noise remains and "music noise" is introduced, and most multi-channel The frequency response compensation algorithm distorts the speech structure, resulting in unsatisfactory speech intelligibility and comfort for the wearers of digital hearing aids. A method of combining speech enhancement and frequency response compensation is proposed. The algorithm makes full use of The working mechanism of the human ear is preserved, the speech features are preserved, and the "music noise" introduced by the spectral subtraction method is removed, so that the intelligibility of the hearing aid output signal is greatly improved; since this algorithm is based on the spectral subtraction method, Therefore, the algorithm is simple to implement, low complexity, and low power consumption

Images