[en] Sound source localization and tracking algorithms scale badly as the number of microphones grows, leading to bottlenecks in real-time implementations. This work presents a novel hybrid algorithm that manages complexity blowup. We combine the Generalized Cross Correlation (GCC) and Delay and Sum Beamforming (DSB) localization algorithms adapting the computation spectrum to the Field of View (FoV). We select the appropriated GCC algorithm according to the FoV and the number of microphones used. We then demonstrate that more than 70% of the DSB computation can be eliminated without affecting the sound source localization.
Disciplines :
Electrical & electronics engineering Library & information sciences
Author, co-author :
Ibala, Serge ; Université de Mons > Faculté Polytechnique > Electronique et Microélectronique
Bettens, Frédéric ; Université de Mons > Faculté Polytechnique > Information, Signal et Intelligence artificielle
Valderrama, Carlos ; Université de Mons > Faculté Polytechnique > Electronique et Microélectronique
Language :
English
Title :
Faster Sound Source Localization and Tracking with Limited Computation Load
Publication date :
04 July 2012
Event name :
The 6th International Symposium on signal, Image, Video and Communications ISIVC 2012
Event place :
Valenciennes, France
Event date :
2012
Research unit :
F105 - Information, Signal et Intelligence artificielle F109 - Electronique et Microélectronique
