Comparing Separation Quality of Nonnegative Matrix Factorization and Nonnegative Matrix Factor 2D Deconvolution in Audio Source Separation Tasks

The Nonnegative Matrix Factorization (NMF) is widely used in audio source separation tasks. However, the separation quality of NMF varies a lot depending on the mixture. In this paper, we analyze the use of NMF in source separation tasks and show how separation results can be significantly improved by using the Nonnegative Matrix Factor 2D Deconvolution (NMF2D). NMF2D was originally proposed as an extension to the NMF to circumvent the problem of grouping notes, but it is used differently in this paper to improve the separation quality, without taking the problem of grouping notes into account.