Wavelet Based Speech Enhancement Algorithm for Hearing Aid Application

There are many factors which causes hearing loss e.g., listening to loud music and excessive noise exposure in daily work environment. Ageing is also another major factor which causes hearing loss. Frequency-dependent hearing loss is the general hearing aid impaired that most of the people are suffering in this world. The frequency dependent and a level-dependent gain are implemented for compensating this kind of hearing loss. Dynamic range compression (DRC) corresponding the digital signal processing (DSP) algorithm is used in this method. The significant problem in hearing impaired person is background noise in which the person cannot understand the speech in noise, hence there is a need for higher signal-to-noise-ratio (SNR). The another important algorithm component in hearing aid is noise reduction. The SNR reduction is the key feature in hearing aid. Thus the output obtained must have low SNR value. A set of speech filtering methods are used to obtain better results to reduce the noise present in speech signal. Implementations are performed in MATLAB. A set of objective experiments and experimental measures are developed to verify the performance of the system.