A Framework for Analysis of Music Similarity Measures

Jesper Højvang Jensen; Mads G. Christensen; Søren Holdt Jensen

A Framework for Analysis of Music Similarity Measures

Jesper Højvang Jensen, Mads G. Christensen, Søren Holdt Jensen

Department of Electronic Systems

Research output: Contribution to book/anthology/report/conference proceeding › Article in proceeding › Research › peer-review

3 Citations (Scopus)

781 Downloads (Pure)

Abstract

To analyze specific properties of music similarity measures that the commonly used genre classification evaluation procedure does not reveal, we introduce a MIDI based test framework for music similarity measures. We introduce the framework by example and thus outline an experiment to analyze the dependency of a music similarity measure on the instrumentation of a song compared to the melody, and to analyze its sensitivity to transpositions.

Using the outlined experiment, we analyze music similarity measures from three software packages, namely Marsyas, MA toolbox and Intelligent Sound Processing toolbox. The tested timbral similarity measures perform instrument recognition relatively well, although they are sensitive to transpositions and differences between sound fonts. The beat/rhythm/melody similarity measures are not always able to recognize the same melody played with different instruments.

Original language	English
Title of host publication	Proceedings of the 15th European Signal Processing Conference
Number of pages	5
Publisher	EURASIP Open Library
Publication date	2007
Pages	926-930
Publication status	Published - 2007
Event	European signal processing conference - Poznań, Poland Duration: 3 Sept 2007 → 7 Sept 2007 Conference number: 15

Conference

Conference	European signal processing conference
Number	15
Country/Territory	Poland
City	Poznań
Period	03/09/2007 → 07/09/2007

Access to Document

Jensen07eusipcoFinal published version, 450 KB

AUB Link

Search for the material in Aalborg University Library's search engine

Intelligent Sound
Jensen, C. S., Pedersen, T. B., Schmidt, A., Deliege, F. & Ruxanda, M. M.
FTP Forskningsrådet for Teknologi og Produktion
01/01/2005 → 31/12/2008
Project: Research

Cite this

@inproceedings{6088e5409c2c11db8ed6000ea68e967b,

title = "A Framework for Analysis of Music Similarity Measures",

abstract = "To analyze specific properties of music similarity measures that the commonly used genre classification evaluation procedure does not reveal, we introduce a MIDI based test framework for music similarity measures. We introduce the framework by example and thus outline an experiment to analyze the dependency of a music similarity measure on the instrumentation of a song compared to the melody, and to analyze its sensitivity to transpositions.Using the outlined experiment, we analyze music similarity measures from three software packages, namely Marsyas, MA toolbox and Intelligent Sound Processing toolbox. The tested timbral similarity measures perform instrument recognition relatively well, although they are sensitive to transpositions and differences between sound fonts. The beat/rhythm/melody similarity measures are not always able to recognize the same melody played with different instruments.",

author = "Jensen, {Jesper H{\o}jvang} and Christensen, {Mads G.} and Jensen, {S{\o}ren Holdt}",

year = "2007",

language = "English",

pages = "926--930",

booktitle = "Proceedings of the 15th European Signal Processing Conference",

publisher = "EURASIP Open Library",

note = "European signal processing conference ; Conference date: 03-09-2007 Through 07-09-2007",

}

TY - GEN

T1 - A Framework for Analysis of Music Similarity Measures

AU - Jensen, Jesper Højvang

AU - Christensen, Mads G.

AU - Jensen, Søren Holdt

N1 - Conference code: 15

PY - 2007

Y1 - 2007

N2 - To analyze specific properties of music similarity measures that the commonly used genre classification evaluation procedure does not reveal, we introduce a MIDI based test framework for music similarity measures. We introduce the framework by example and thus outline an experiment to analyze the dependency of a music similarity measure on the instrumentation of a song compared to the melody, and to analyze its sensitivity to transpositions.Using the outlined experiment, we analyze music similarity measures from three software packages, namely Marsyas, MA toolbox and Intelligent Sound Processing toolbox. The tested timbral similarity measures perform instrument recognition relatively well, although they are sensitive to transpositions and differences between sound fonts. The beat/rhythm/melody similarity measures are not always able to recognize the same melody played with different instruments.

AB - To analyze specific properties of music similarity measures that the commonly used genre classification evaluation procedure does not reveal, we introduce a MIDI based test framework for music similarity measures. We introduce the framework by example and thus outline an experiment to analyze the dependency of a music similarity measure on the instrumentation of a song compared to the melody, and to analyze its sensitivity to transpositions.Using the outlined experiment, we analyze music similarity measures from three software packages, namely Marsyas, MA toolbox and Intelligent Sound Processing toolbox. The tested timbral similarity measures perform instrument recognition relatively well, although they are sensitive to transpositions and differences between sound fonts. The beat/rhythm/melody similarity measures are not always able to recognize the same melody played with different instruments.

M3 - Article in proceeding

SP - 926

EP - 930

BT - Proceedings of the 15th European Signal Processing Conference

PB - EURASIP Open Library

T2 - European signal processing conference

Y2 - 3 September 2007 through 7 September 2007

ER -

A Framework for Analysis of Music Similarity Measures

Abstract

Conference

Access to Document

AUB Link

Fingerprint

Projects

Intelligent Sound

Cite this