Get rid of the noise! : A speech enhancement technology using microphone arrays

Abstract

Blind source separation (BSS) is the process that the speech signals of multiple speakers are extracted from a given mixture, with little or no information about the sources and the mixing process. This project is aiming at designing and developing an application on Android devices for implementing BSS. The algorithm this application uses is directional sparse filtering (DSF), based on one of the most popular BSS methods, sparse component analysis (SCA). Compared to the existing application with similar functions, this application shows significant efficiency improvement and separated audio quality enhancement.