Title :
Reduced-rank modeling of head-related impulse responses using subset selection
Author_Institution :
Wave Arts Inc., Arlington, MA
Abstract :
This paper describes the subset selection method of modeling head-related impulse responses. The subset method is mathematical technique for choosing a set of basis functions directly from the data set that is being modeled. The subset method is compared to principal component analysis and the virtual speaker method
Keywords :
audio signal processing; frequency synthesizers; hearing; loudspeakers; transfer functions; transient response; HRTF; basis functions; data set; head-related impulse responses; mathematical technique; reduced-rank modeling; subset selection; Bismuth; Convolution; Eigenvalues and eigenfunctions; Frequency domain analysis; Loudspeakers; Mathematical model; Matrix decomposition; Principal component analysis; Singular value decomposition; Transfer functions;
Conference_Titel :
Applications of Signal Processing to Audio and Acoustics, 1999 IEEE Workshop on
Conference_Location :
New Paltz, NY
Print_ISBN :
0-7803-5612-8
DOI :
10.1109/ASPAA.1999.810878
