Fusion Algorithm Based on Speech Enhancement and Loudness Compensation for Digital Hearing Aids

Ru-Wei Li,Long-Tao Cao,Xiao-Qun Yi,Shuai Wang
DOI: https://doi.org/10.1142/9789814719810_0096
2016-01-01
Abstract:In order to solve the residual background noise resulted by the existing speech enhancement algorithm, and distorting formant of speech resulted by the existing loudness compensation algorithm for hearing aids, the fusion algorithm based on speech enhancement and loudness compensation for digital hearing aids is proposed in this paper. According to human auditory perception theory, noisy is divided into several sub-bands by Gammatone filter banks simulated the cochlear model in this algorithm. In each sub-band denoising and loudness compensation are achieved by multi-masking and sound pressure gain. Experimental results show that the background noise is removed effectively, the requirements of people who have hearing loss of sound pressure level are met, and formant of speech get well protected. Clinical measures show that the intelligibility, articulation and comfort are improved for hearing losses of the patients.
What problem does this paper attempt to address?