Multi-channel noise reduction for hands-free voice communication on mobile phones

Noise reduction technologies have been applied to enhance the intelligibility of voice communications. However, existing methods are vulnerable to complex non-stationary noisy conditions, which are commonly encountered in real world hands-free scenarios. Additionally, the existing methods do not fully take the advantage of the deployment of multi-channel microphone arrays on the burgeoning high-end mobile phones. In this paper, we propose a noise estimation method based on a hybrid of efficient single channel and globally optimized multi-channel microphone noise estimation using adaptive noise coherence models. Evaluation results on real-world recordings collected via a smartphone confirm its superior effectiveness to suppress the fast-varying noises compared to the state-of-the-art baseline methods.