Identifying Saxophonists from Their Playing Styles - Intelligent Music Information Systems: Tools and Methodologies - page 118

our saxophone separation method works only
for a small subset of the polyphonic multi-intru-
ment audio recordings. It remains a challenge
the implementation of a more general method for
obtaining the melody played by the saxophone
from such recordings.

acknoWledgments

This work was supported by the Spanish Minis-
try of Science and Education project ProSeMus
(TIN2006-14932-C02-01). We would like to thank
Esteban Maestre, Emilia Gomez, and Maarten
Grachten for processing the data.

conclusIon

references

In this chapter we focused on the task of identi-
fying performers from their playing style using
note descriptors extracted from audio recordings.
In particular, we concentrated in identifying Jazz
saxophonists and explored and compared differ-
ent machine learning techniques for this task.
We characterized performances by representing
each note in the performance by a set of percep-
tual features corresponding to the perceptual
features of the note, and a set of contextual fea-
tures representing the context in which the note
appears. We presented successful classifiers for
identifying saxophonists in recordings obtained
in a controlled environment. The results obtained
indicate that the perceptual and contextual fea-
tures presented contain sufficient information
to identify the studied set of interpreters, and
that the machine learning methods explored are
capable of learning performance patterns that
distinguish these interpreters. We are currently
extending our approach to performance-based
interpreter identification in polyphonic multi-
instrument audio recordings with encouraging
preliminary results. We intend to pursue this
research line. All in all, given the capabilities of
current audio analysis systems obtained results,
we believe expressive-content-based performer
identification is a promising research topic in
music information retrieval.

Apel, W. (1972). Harvard dictionary of music (2nd
ed.). Cambridge, MA: Belknap Press of Harvard
University Press.

Bernstein, A. D., & Cooper E. D. (1976, July/Au-
gust). The piecewise-linear technique of electronic
music synthesis. J. Audio Eng. Soc., 24 (6).

Bresin, R. (2002). Articulation rules for automatic
music performance. In Proceedings of the 2001
International Computer Music Conference , San
Francisco, California.

Canazza, S., De Poli, G., Roda, A., & Vidolin, A.
(1997). Analysis and synthesis of expressive inten-
tion in a clarinet performance. In Proceedings of
the 1997 International Computer Music Confer-
ence (pp. 113-120). San Francisco, California.

Cano, P. (1998). Fundamental frequency estima-
tion in the SMS analysis. In Proceedings of the
Digital AudioEffects Workshop (DAFx) , Barce-
lona, Spain.

Chauvin, Y. et al. (1995). Backpropagation:
Theory, architectures and applications . Lawrence
Erlbaum Assoc.

Colmenauer A. (1990). An introduction to PRO-
LOG-III. Communications of the ACM, 33 (7).