Music signal separation by supervised nonnegative matrix factorization with basis deformation

In this paper, we address a music signal separation problem, and propose a new supervised algorithm for real instrumental signal separation employing a deformable capability for a spectral supervision trained in advance. Nonnegative matrix factorization (NMF) is one of the techniques used for the separation of an audio mixture that consists of multiple instrumental sources. Conventional supervised NMF has the critical problem that a mismatch between the bases trained in advance and the target real sound reduces the accuracy of separation. To solve this problem, we propose a new advanced supervised NMF that employs a deformable capability for the trained bases and penalty terms for making the bases fit into the target sound. The results of the experiment using real instruments show that the proposed method significantly improves the accuracy of separation compared with the conventional method.

I. Introduction

In recent years, source separation based on nonnegative matrix factorization (NMF) [1], which is a type of sparse representation algorithm, has been a very active area of signal processing research. NMF for acoustical signals decomposes an input spectrogram into a product of a spectral basis matrix and its activation matrix. In particular, NMF has been a convincing candidate for source separation in musical signal processing with a monaural format [2].