ISMIR 2005
6th International Conference on Music Information Retrieval

London, UK        11 - 15 September 2005

Tutorials

The ISMIR 2005 Tutorials will take place on Sunday 11 September 2005 from 1:30pm-6:00pm in the Lecture Theatre, Ben Pimlott Building, Goldsmiths College, University of London.

Tutorial 1: Music Background

Part I: The combination of musicology, engineering, and interpretation in Beethoven's 5th symphony
Prof. Dalia Atlas, Guest conductor of 71 major orchestras in 30 countries, Repertoire of over 750 scores.

Part II: Looking Beyond the Notes: An Engineering Approach to the Study of Musical Performance
Prof. Elaine Chew, Integrated Media Systems Center, University Southern California

Abstract: Designed for professionals and non-specialists, the first part of this tutorial will present an analysis of the Fifth Symphony by Beethoven, revealing as many aspects as possible, appropriate for professional musicians and amateurs at the same time. The second part will consist of a detailed look at work on measuring the performance characteristics of musicians using new technologies.

Tutorial 2: Music Similarity and Visualization for Music IR

Part I: Music Similarity
Elias Pampalk, ÖFAI (Austrian Research Institute for Artificial Intelligence)

Part II: Visualization Methods for MIR
Stephan Baumann, DFKI GmbH (German Research Center for AI)

Abstract: The first part of this tutorial is an introduction to the computation of audio and web-based music similarity. This tutorial will cover low-level audio statistics related to timbre and rhythm as well as the application of text information retrieval techniques to MIR. In particular, the algorithm which won the ISMIR'04 genre classification contest will be described. We illustrate the use of these techniques for playlist generation and genre classification. The UI and visualization part of this tutorial will look at 2D methods, 3D methods, data mining visualization, animation and interfaces for inspecting and navigating the music information based on the similarity computations.