A reconfigurable sound wave decomposition filterbank for hearing aids based on nonlinear transformation


Hearing impaired people have their own hearing loss characteristics and listening preferences. Therefore hearing aid system should become more natural, humanized and personalized, which requires the filterbank in hearing aids provides flexible sound wave decomposition schemes, so that patients are likely to use the most suitable scheme for their own hearing compensation strategy. In this paper, a reconfigurable sound wave decomposition filterbank is proposed. The prototype filter is first cosine modulated to generate uniform subbands. Then by non-linear transformation the uniform subbands are mapped to nonuniform subbands. By changing the control parameters, the nonlinear transformation changes which leads to different subbands allocations. It provides four different sound wave decomposition schemes without changing the structure of the filterbank. The performance of the proposed reconfigurable filterbank was compared with that of fixed filerbanks, fully customizable filterbanks and other existing reconfigurable filterbanks. It is shown that the proposed filterbank provides satisfactory matching performance as well as low complexity and delay, which make it suitable for real hearing aid applications.

Shaoguang Huang
Postdoctoral researcher

My research interests include image processing, sparse modelling and hyperspectral image analysis.