Non-negative matrix factorization for polyphonic music transcription
Authors:
- P. SmaragdisPSP. Smaragdis
- J.C. BrownJBJ.C. Brown
Top Cited Papers
- 6 May 2004
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 177-180
Abstract
In this paper we present a methodology for analyzing polyphonic musical passages comprised by notes that exhibit a harmonically fixed spectral profile (such as piano notes). Taking advantage of this unique note structure we can model the audio content of the musical passage by a linear basis transform and use non-negative matrix decomposition methods to estimate the spectral profile and the temporal information of every note. This approach results in a very simple and compact system that is not knowledge-based, but rather learns notes by observation.Keywords
This publication has 5 references indexed in Scilit:
- Non-negative sparse coding2003
- Automatic Music Transcription and Audio Source SeparationCybernetics and Systems, 2002
- Conditions for nonnegative independent component analysisIEEE Signal Processing Letters, 2002
- Learning the parts of objects by non-negative matrix factorizationNature, 1999
- Least squares formulation of robust non-negative factor analysisChemometrics and Intelligent Laboratory Systems, 1997