Improving the Classification of Percussive Sounds with Analytical Features: A Case Study

By Pierre Roy, François Pachet, and Sergio Krakowski.

Proceedings of the 8th International Conference on Music Information Retrieval (ISMIR 2007), pp. 229–232, Vienna, Austria, 2007.

Read the paper (PDF)
ISMIR proceedings PDF

About the paper

This paper tests automatically constructed audio features on pandeiro sounds. Pierre Roy, François Pachet and Sergio Krakowski compare analytical features with standard descriptors and features designed by hand for the task.

Abstract summary

The study considers two problems: classifying complete percussive sounds for retrieval, and classifying their attack portions for interactive music applications. The system constructs, evaluates and selects compositions of signal-processing operators. The experiments measure the improvement that these task-specific features bring over the comparison methods.

Related version

Analytical Features for the Classification of Percussive Sounds: The Case of the Pandeiro (DAFx 2007 PDF) — related work by the same three authors.

Related work

Feature Generation project
Automatic Extraction of Music Descriptors from Acoustic Signals Using EDS
Analytical Features: A Knowledge-Based Approach to Audio Feature Generation

Back to Publications