Audio Source Positioning Based on Angle of Arrival Measurements Using Particle Filter

Abstract

This paper presents a numerical approach to the audio source positioning problem using microphones available in mobile devices. The proposed solution is a particle filter. The algorithm accuracy was tested with four different datasets, two measured in an anechoic chamber and two in a listening room. The particle filters smoothing parameter and measurement noise variance are estimated offline. The results show that this algorithm is able to position the audio source well in quiet environments.