TY - GEN
T1 - Single channel source separation using smooth Nonnegative Matrix Factorization with Markov Random Fields
AU - Kim, Minje
AU - Smaragdis, Paris
PY - 2013
Y1 - 2013
N2 - This paper presents a single channel source separation method based on an extension of Nonnegative Matrix Factorization (NMF) algorithm by smoothing the original posterior probabilities with an additional Markov Random Fields (MRF) structure. Our method is based on the alternative interpretation of NMF with β-divergence as latent variable models. By doing so, we can redefine NMF-based separation procedure as a Bayesian labeling problem where each label stands for the mask for a specific source. This understanding leads us to intervene in the calculation of posterior probabilities, so that the priors from MRF's neighboring structure can smooth out isolated masking values that have different labeling results from their neighbors. Experiments on several dictionary-based source separation tasks show sensible performance gains.
AB - This paper presents a single channel source separation method based on an extension of Nonnegative Matrix Factorization (NMF) algorithm by smoothing the original posterior probabilities with an additional Markov Random Fields (MRF) structure. Our method is based on the alternative interpretation of NMF with β-divergence as latent variable models. By doing so, we can redefine NMF-based separation procedure as a Bayesian labeling problem where each label stands for the mask for a specific source. This understanding leads us to intervene in the calculation of posterior probabilities, so that the priors from MRF's neighboring structure can smooth out isolated masking values that have different labeling results from their neighbors. Experiments on several dictionary-based source separation tasks show sensible performance gains.
KW - Informed Source Separation
KW - Markov Random Fields
KW - Nonnegative Matrix Factorization
KW - Probabilistic Latent Component Analysis
KW - Probabilistic Latent Semantic Indexing
UR - http://www.scopus.com/inward/record.url?scp=84893338433&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84893338433&partnerID=8YFLogxK
U2 - 10.1109/MLSP.2013.6661909
DO - 10.1109/MLSP.2013.6661909
M3 - Conference contribution
AN - SCOPUS:84893338433
SN - 9781479911806
T3 - IEEE International Workshop on Machine Learning for Signal Processing, MLSP
BT - 2013 IEEE International Workshop on Machine Learning for Signal Processing - Proceedings of MLSP 2013
T2 - 2013 16th IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2013
Y2 - 22 September 2013 through 25 September 2013
ER -