Improving Melody Classification by Discriminant Feature Extraction and Fusion

M. Li; M. R. Sleep

Improving Melody Classification by Discriminant Feature Extraction and Fusion

School of Computing Sciences

Research output: Contribution to conference › Paper

Abstract

We propose a general approach to discriminant feature extraction and fusion, built on an optimal feature transformation for discriminant analysis [6]. Our experiments indicate that our approach can dramatically reduce the dimensionality of original feature space whilst improving its discriminant power. Our feature fusion method can be carried out in the reduced lower-dimensional subspace, resulting in a further improvement in accuracy. Our experiments concern the classification of music styles based only on the pitch sequence derived from monophonic melodies

Original language	English
Publication status	Published - Oct 2004
Event	ISMIR2004, 5th International Conference on Music Information Retrieval - Barcelona, Spain Duration: 10 Oct 2004 → 14 Oct 2004

Conference

Conference	ISMIR2004, 5th International Conference on Music Information Retrieval
Country/Territory	Spain
City	Barcelona
Period	10/10/04 → 14/10/04

Cite this

@conference{21271fc6551a44b698f150b1b1000590,

title = "Improving Melody Classification by Discriminant Feature Extraction and Fusion",

abstract = "We propose a general approach to discriminant feature extraction and fusion, built on an optimal feature transformation for discriminant analysis [6]. Our experiments indicate that our approach can dramatically reduce the dimensionality of original feature space whilst improving its discriminant power. Our feature fusion method can be carried out in the reduced lower-dimensional subspace, resulting in a further improvement in accuracy. Our experiments concern the classification of music styles based only on the pitch sequence derived from monophonic melodies",

author = "M. Li and Sleep, {M. R.}",

year = "2004",

month = oct,

language = "English",

note = "ISMIR2004, 5th International Conference on Music Information Retrieval ; Conference date: 10-10-2004 Through 14-10-2004",

}

TY - CONF

T1 - Improving Melody Classification by Discriminant Feature Extraction and Fusion

AU - Li, M.

AU - Sleep, M. R.

PY - 2004/10

Y1 - 2004/10

N2 - We propose a general approach to discriminant feature extraction and fusion, built on an optimal feature transformation for discriminant analysis [6]. Our experiments indicate that our approach can dramatically reduce the dimensionality of original feature space whilst improving its discriminant power. Our feature fusion method can be carried out in the reduced lower-dimensional subspace, resulting in a further improvement in accuracy. Our experiments concern the classification of music styles based only on the pitch sequence derived from monophonic melodies

AB - We propose a general approach to discriminant feature extraction and fusion, built on an optimal feature transformation for discriminant analysis [6]. Our experiments indicate that our approach can dramatically reduce the dimensionality of original feature space whilst improving its discriminant power. Our feature fusion method can be carried out in the reduced lower-dimensional subspace, resulting in a further improvement in accuracy. Our experiments concern the classification of music styles based only on the pitch sequence derived from monophonic melodies

M3 - Paper

T2 - ISMIR2004, 5th International Conference on Music Information Retrieval

Y2 - 10 October 2004 through 14 October 2004

ER -