conference paper

Conference/ProceedingsInternational Conference on Music Information Retrieval (ISMIR 2007)
Start date23.09.2007
End date27.09.2007
AddressVienna, Austria
Author(s)Luis Gustavo Martins, Juan José Burred, George Tzanetakis, Mathieu Lagrange
TitlePolyphonic Instrument Recognition Using Spectral Clustering
AbstractThe identification of the instruments playing in a polyphonic music signal is an important and unsolved problem in Music Information Retrieval. In this paper, we propose a framework for the sound source separation and timbre classification of polyphonic, multi-instrumental music signals. The sound source separation method is inspired by ideas from Computational Auditory Scene Analysis and formulated as a graph partitioning problem. It utilizes a sinusoidal analysis front-end and makes use of the normalized cut, applied as a global criterion for segmenting graphs. Timbre models for six musical instruments are used for the classification of the resulting sound sources. The proposed framework is evaluated on a dataset consisting of mixtures of a variable number of simultaneous pitches and instruments, up to a maximum of four concurrent notes. The overall instrument classification success rate is of 47%.
NoteL.G. Martins: INESC Porto, Portugal
G. Tzanetakis, M. Lagrange: University of Victoria, Canada