< Back to previous page

Project

Signal processing algorithms for binaural hearing aids.

Modern digital hearing aids aren't simple sound amplifiers, but also
contain signal processing algorithms for noise reduction, feedback
suppression,... Most hearing aid users have hearing aids at both ears
in a so-called bilateral configuration, which means the hearing aids
work independently.
The independent processing of the bilateral configuration destroys the
so-called binaural cues (the time and level difference of the received
signals at both ears), which are used by us to localise sounds. A
binaural algorithm, where the hearing aids can communicate over a
wireless link, should be able to coordinate its signal processing so
that the binaural cues are preserved. A correct localisation can help
in speech understanding, by the effect of spatial unmasking.
Our goal is therefore to develop a binaural noise reduction algorithm
which also preserves binaural cues. Existing monaural noise reduction
algorithms based on Multi-channel Wiener Filtering will be extended,
while taking practical constraints into account: low power
consumption, limited bandwidth of the wireless link and small
input-output delay.
Date:1 Jan 2008 →  31 Dec 2011
Keywords:Binaural hearing aids
Disciplines:Applied mathematics in specific fields
Project type:PhD project