Statistical description models for melody analysis and characterization
Pedro J. Ponce de León and José Manuel Iñesta
In: Proceedings of the International Computer Music Conference (2004) Jet Graphics and Mailers , Miami, USA , pp. 149-156. ISBN 0-9713192-2-7

Abstract

The analysis of melodies symbolically represented as digital scores (standard MIDI files) is studied. A number of melodic, harmonic, and rhythmic statistical descriptors are computed. Their representation ability is analyzed using the principal components technique (PCA). Also their ability to separate two particular musical styles, like jazz and classical, using a per-feature separability test. A qualitative discussion on such a description model based on these results is included. {\bf Keywords: Melody description, principal component analysis, music information retrieval.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.

[Edit]