Source/filter model for unsupervised main melody extraction from polyphonic audio signals (2010)

by J-L Durrieu, G Richard, B David, C Févotte
Venue:IEEE Transactions on Audio, Speech and Language Processing