Single Channel Speech Music Separation Using Nonnegative Matrix Factorization with Sliding Windows and Spectral Masks.

Emad M. Grais, Hakan Erdogan

Browse the full Interspeech paper archive.