BIOTOP: Adaptive Wavelet and Frame techniques for acoustic BEM. Boundary Integral Operator Solution Techniques with Optimal Properties
Main Applicant: Wolfgang Kreuzer (Austrian Academy of Sciences, Acoustics Research Institute)
Co-Applicants: Peter Balazs (Austrian Academy of Sciences, Acoustics Research Institute), Stephan Dahlke (Phillipps-University Marburg) and Helmut Harbrecht (University Basel)
Start of the Project: 01. May 2013
The projects aims at developing efficient methods to calculate head related transfer functions (HRTFs). HRTFs describe the acoustic filtering effect of pinna (outer ear), head and torso on incoming sound. Thus, they are fundamental for sound localization in humans, for example the detection of an approaching car. For acoustic simulations the boundary element method (BEM) is a commonly used tool. However, the BEM has the big drawback that the computational effort and hardware requirement grows with the frequency. As HRTFs need to be calculated for frequencies up to 20.000 Hertz the BEM could only be used partially for HRTF calculations in the past. The efficient algorithms developed in BIOTOP shall help to deal with this drawback and make way for an efficient calculation of HRTFs in the human audio frequency range.
In BIOTOP the number of calculations shall be reduced by using adaptive wavelet- and frame techniques, thus making the BEM feasable for HRTF-calculations. Compared to commonly used BEM basis functions, wavelets have the advantage that a wavelet transformation provides functions that can adapt better to a given distribution of the acoustic field on the head. As a generalization of wavelets, frames allow for an even more flexible construction method and thus for a better adaption to the problem at hand.