Multi-source separation using over iterative empirical mode decomposition

Abstract

In this paper, we propose a novel multi-source fundamental frequency estimation method for music signals which is based on over iterative empirical mode decomposition (EMD). Multi-source fundamental frequency estimation in single channel polyphonic signal has been developed for a long time but still a challenging problem. By means of using over iterative EMD, we can obtain predominant frequency modulation subband signal named Intrinsic Mode Function (IMF) without the interference of amplitude modulation. Then the fundamental frequencies are obtained by a salience estimation method. The experimental results agree that the proposed method can extract the pitch content efficiently. © 2013 Springer-Verlag.