A New Approach for Enhancing MVDR Beamformer’s Performance

Abstract

Minimum Variance Distortionless Response (MVDR) has many applications in the microphone array processing approach due to its advantages of suppressing background noise, interference, or third-party speaker. However, in real-life recording situations, the performance of the MVDR beamformer is often degraded by microphone mismatches, different microphone sensitivities, or errors in microphone distribution. Therefore, in this paper, the authors suggested a new approach, which uses an additive matrix for modifying the covariance matrix of observed microphone array data. The illustrated experiments were evaluated in a living room with a dual-microphone system (DMA2) and a stand speaker in the presence of noise. The theoretical results are further confirmed in terms of the signal-to-noise ratio (SNR) and the speech intelligibility of the output signal. The proposed method can be integrated into systems with multiple microphones and in annoying complex environments.