An Experimental Survey on Non-Negative Matrix Factorization for Single Channel Blind Source Separation

In applications such as speech and audio denoising, music tran-scription, music and audio based forensics, it is desirable to de-compose a single-channel recording into its respective sources, commonly referred to as blind source separation (BSS). One of the techniques used in BSS is non-negative matrix factorization (NMF). In NMF both supervised and unsupervised mode of op-erations is used. Among them supervised mode outperforms well due to the use of pre-learned basis vectors corresponding to each underlying sources. In this paper NMF algorithms such as Lee Se-ung algorithms (Regularized Expectation Minimization Maximum Likelihood Algorithm (EMML) and Regularized Image Space Re-construction Algorithm (ISRA)), Bregman Divergence algorithm (Itakura Saito NMF algorithm (IS-NMF)) and an extension to NMF, by incorporating sparsity, Sparse Non-Negative Matrix Fac-torization(SNMF) algorithm are used to evaluate the performance of BSS in which supervised mode is used. Here signal to distor-tion ratio (SDR), signal to interference ratio (SIR) and signal to artifact ratio (SAR) are measured for different speech and/or mu-sic mixtures and performance is evaluated for each combination. General Terms