Performance analysis of the extended binaural MVDR beamformer with partial noise estimation in a homogeneous noise field
Nico Gößling, Daniel Marquardt, Simon Doclo · 2017
Besides noise reduction an important objective of a binaural speech enhancement algorithm is the preservation of the binaural cues of both the desired speech source as well as the undesired noise sources. Recently, it has been shown that by combining a binaural hearing system with an external microphone it is possible to improve both the noise reduction performance as well as the binaural cue preservation for a directional noise source. In this paper, we consider the binaural minimum variance distortionless response beamformer with partial noise estimation (MVDR-N), which aims at preserving the speech component and a scaled version of the noise component. For a homogeneous noise field, we theoretically analyze the influence of an external microphone on the performance of the binaural MVDR-N beamformer, showing that a larger SNR improvement can be achieved and a smaller scaling parameter is required to achieve a desired interaural coherence for the residual noise. Experimental simulation results in a realistic acoustic scenario confirm these theoretical findings.