Research On Key Technologies Of Digital Hearing AIDS For The Elderly

In today's fast-paced and high-intensity work pace,the number of people suffering from hearing impairment of varying degrees is increasing year by year,whether due to congenital physical defects or acquired environmental influences.As a member of the society,communication is the most effective way to integrate into the society.However,for hearing impaired people,hearing AIDS help them to integrate into the society.But because of the small size of digital hearing aid,microphone and speaker near the table,resulting in digital hearing aid often whistles,signal distortion.Therefore,digital hearing aid technology for hearing impairment needs further research.On the basis of studying the traditional algorithm and summarizing the previous algorithms,this paper proposes the improved algorithm of speech denoising and speech enhancement.The main work is as follows:1.The basic knowledge of speech signal generation of digital hearing AIDS is introduced,including the generation principle and basic properties of speech signal,three models of speech signal generation and the analysis method of speech signal;The key technologies of digital hearing AIDS are studied,including speech enhancement technology,sound source location technology,wide dynamic compression technology and echo feedback elimination.Focus on Speech enhancement technology2.The adaptive beamforming algorithm of generalized sidelobe canceller is studied.Based on the analysis of speech leakage,a wavelet threshold denoising algorithm is introduced,and an adaptive beamforming algorithm based on wavelet denoising is proposed.The advantage of wavelet denoising is that it can enhance the signal-to-noise ratio of the input signal and make the speech output after passing the GSC structure clearer and more accurate.Through further simulation,the results show that the algorithm can solve the problem of speech leakage and residual noise,and improve the recognition of speech.3.The adaptive echo cancellation algorithm of digital hearing AIDS is studied,and the common LMS and NLMS are derived.The two algorithms were simulated and analyzed by using the recorded audio in Matlab software,and the results showed that the NLMS algorithm was more in line with the adaptive echo cancellation.