Two microphone approach for multiple sound source tracking
Omiya Kenta, Kenji Suyama · 2016
In this paper, a method for multiple sound source tracking using two microphones is proposed. In our previous method, a penalty function for PSO (Particle Swarm Optimization) was added for emphasizing correct peaks on the MUSIC (MUltiple SIgnal Classification) spectrum. However, the penalty function was fixed regardless of the tracking situation. In the proposed method, the suitable penalty function based on the denominator of MUSIC spectrum is generated depending on the tracking situation. The effectiveness of the proposed method is shown by several experimental results.