文档介绍:Digital Signal Processing 29 (2014) 20–34
Contents lists available at ScienceDirect
Digital Signal Processing
Source separation using regularized NMF with MMSE estimates under
GMM priors with online learning for the uncertainties
∗
Emad M. Grais , Hakan Erdogan
Faculty of Engineering and Natural Sciences, Sabanci University, Orhanli Tuzla, 34956, Istanbul, Turkey
article info abstract
Article history: We propose a new method to incorporate priors on the solution of nonnegative matrix factorization
Available online 12 March 2014 (NMF). The NMF solution is guided to follow the minimum mean square error (MMSE) estimates of the
weight combinations under a Gaussian mixture model (GMM) prior. The proposed algorithm can be used
Keywords:
for denoising or single-channel source separation (SCSS) applications. NMF is used in SCSS in two main
Single channel source separation
stages, the training stage and the separation stage. In the training stage, NMF is used to decompose the
Nonnegative matrix factorization
Minimum mean square error estimates training data spectrogram for each source into a multiplication of a trained basis and gains matrices. In
Gaussian mixture models the separation stage, the mixed signal spectrogram is decomposed as a weighted linear combination of
the trained basis matrices for the source signals. In this work, to improve the separation performance of
NMF, the tra