Music Signal Separation Based on Supervised Nonnegative Matrix Factorization with Orthogonality and Maximum-Divergence Penalties

Daichi KITAMURA; Hiroshi SARUWATARI; Kosuke YAGI; Kiyohiro SHIKANO; Yu TAKAHASHI; Kazunobu KONDO

doi:10.1587/transfun.E97.A.1113

Regular Section

Music Signal Separation Based on Supervised Nonnegative Matrix Factorization with Orthogonality and Maximum-Divergence Penalties

Daichi KITAMURA, Hiroshi SARUWATARI, Kosuke YAGI, Kiyohiro SHIKANO, Yu TAKAHASHI, Kazunobu KONDO

Author information

Keywords: music signal separation, nonnegative matrix factorization, supervised method

JOURNAL RESTRICTED ACCESS

2014 Volume E97.A Issue 5 Pages 1113-1118

DOI https://doi.org/10.1587/transfun.E97.A.1113

Details

Abstract

In this letter, we address monaural source separation based on supervised nonnegative matrix factorization (SNMF) and propose a new penalized SNMF. Conventional SNMF often degrades the separation performance owing to the basis-sharing problem. Our penalized SNMF forces nontarget bases to become different from the target bases, which increases the separated sound quality.

Corresponding author

Register with J-STAGE for free!