You are here

Multi-channel source enhancement

Given a multi-microphone input, several possible techniques can be applied in noisy environments in order to improve the quality of the desired speech signal before feeding a speech recognition or a speech communication chain. Source separation and speech enhancement techniques are commonly used to separate the signal contributions of multiple competing sources and to enhance a specific target source, thus reducing the noise components from noisy mixtures.

Related keywords: microphone arrays, adaptive beamforming, noise reduction, post-filtering, blind source separation, source extraction, acoustic features for robust speech recognition.