Sound source localization using microphone arrays is largely applied in industrial engineering to diagnose acoustic issues appearing in complex machines. Over the past 30 years, a large number of methods have been developed to improve the source identification procedure.
Bayesian Focusing can generalize these methods by automatically calculating the optimal functions based on the geometrical set-up. In addition, probabilistic acoustic inverse problem solving offers the possibility to introduce prior knowledge of the source, noise and propagation model.
MicrodB is a precursor in this technology thanks to its collaboration with the INSA Lyon
The use of Bayesian Focusing capabilities to extract source characteristics (directivity, coherence, sound power) greatly improves sound source identification in industrial applications.
This will be documented at the ISMA conference in Leuven on September 17th to 19th.