Expansion of sound source emphasis to multiple areas

An expansion is introduced to a technique for emphasising sound signals from one particular area that is surrounded by undesired noise sources. The proposed expansion allows sound signals to be extracted from areas located in different positions. The previous method uses multiple fixed beamformers to estimate the power spectral densities of two-dimensional areas and thus, calculate the necessary gain coefficients used to emphasise sources within the target region. The proposed expansion adjusts the beamformer directivity, in order to change the location of the target region. Validation was achieved by testing the performance within an anechoic chamber using two Kinect sensors equipped with microphone arrays. The expansion achieved noise attenuation and signal distortion performance equivalent to that of the previous method in different target regions.