A selective sound receiving microphone system using Blind Source Separation

摘要

This paper describes a new microphone system using two microphones and Blind SourceSeparation. The system can separate objective sound and non-objective sound of thesame direction, and output them in real-time. There is no degradation of tonal qualitydue to signal processing. When the sound and noise are mutually independent and themicrophones pick up a linear combination of the two, we can estimate the mixing matrixcoefftcients by minimizing the cross-correlation of output signals.