Method, apparatus, and manufacture of adaptive null beamforming for a two-microphone array

摘要

Audio signals from a two microphone array (102, fig. 1) are adaptively null-beamformed in order to separate a target signal (eg. speech from noise) by decomposing the signals into subbands and adaptively updating the complex beamforming weights based on the direction θ and distance of the target signal. The step size may be updated proportionally to the target signal power ratio, and a degradation factor β may be based on the source distance.