SOUND SEPARATION USING THE METHOD OF INDEPENDENT COMPONENT ANALYSIS

Abstract

Our ears often simultaneously receive various sound sources (speech, music, noise ...), but we can still listen to the intended sound. A system of speech processing or recognizing must achieve the same intelligent level. The problem is that we receive many mixed (combined) signals from many different source signals, and would like to recover the source signals separately. In the last decade or so a method has been developed to help solve the above problem of Blind Source Separation (BSS) effectively, that is the Independent Component Analysis (ICA). This paper gives an brief overview of ICA and our application to sound separation. The results are very good. Henceforth, we will be able to tackle more complex source separation problems using ICA.