Efficient Similarity Retrieval in Music Databases

Maria Magdalena Ruxanda, Christian Søndergaard Jensen

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

Abstract

Audio music is increasingly becoming available in digital form, and the digital music collections of individuals continue to grow. Addressing the need for effective means of retrieving music from such collections, this paper proposes new techniques for content-based similarity search. Each music object is modeled as a time sequence of high-dimensional feature vectors, and dynamic time warping (DTW) is used as the similarity measure. To accomplish this, the paper extends techniques for time-series-length reduction and lower bounding of DTW distance to the multi-dimensional case. Further, the Vector Approximation file is adapted to the indexing of time sequences and to use a lower bound on the DTW distance. Using these techniques, the paper exploits the lack of a ground truth for queries to efficiently compute query results that differ only slightly from results that may be more accurate, but also are much more expensive, to compute. In particular, the paper demonstrates that aggressive use of time-series length reduction together with query expansion results in significant performance improvements while providing good, approximative query results.
Original languageEnglish
Title of host publicationProceedings of the 13th International Conference on Management of Data
Number of pages12
Publication date2006
Pages56-67
Publication statusPublished - 2006
EventInternational Conference on Management of Data - New Delhi, India
Duration: 14 Dec 200616 Dec 2006
Conference number: 13

Conference

ConferenceInternational Conference on Management of Data
Number13
CountryIndia
CityNew Delhi
Period14/12/200616/12/2006

Fingerprint

Time series

Keywords

  • music similarity retrieval, time series, indexing

Cite this

Ruxanda, M. M., & Jensen, C. S. (2006). Efficient Similarity Retrieval in Music Databases. In Proceedings of the 13th International Conference on Management of Data (pp. 56-67)
Ruxanda, Maria Magdalena ; Jensen, Christian Søndergaard. / Efficient Similarity Retrieval in Music Databases. Proceedings of the 13th International Conference on Management of Data. 2006. pp. 56-67
@inproceedings{a3b878c0a89911dbb942000ea68e967b,
title = "Efficient Similarity Retrieval in Music Databases",
abstract = "Audio music is increasingly becoming available in digital form, and the digital music collections of individuals continue to grow. Addressing the need for effective means of retrieving music from such collections, this paper proposes new techniques for content-based similarity search. Each music object is modeled as a time sequence of high-dimensional feature vectors, and dynamic time warping (DTW) is used as the similarity measure. To accomplish this, the paper extends techniques for time-series-length reduction and lower bounding of DTW distance to the multi-dimensional case. Further, the Vector Approximation file is adapted to the indexing of time sequences and to use a lower bound on the DTW distance. Using these techniques, the paper exploits the lack of a ground truth for queries to efficiently compute query results that differ only slightly from results that may be more accurate, but also are much more expensive, to compute. In particular, the paper demonstrates that aggressive use of time-series length reduction together with query expansion results in significant performance improvements while providing good, approximative query results.",
keywords = "music similarity retrieval, time series, indexing",
author = "Ruxanda, {Maria Magdalena} and Jensen, {Christian S{\o}ndergaard}",
year = "2006",
language = "English",
pages = "56--67",
booktitle = "Proceedings of the 13th International Conference on Management of Data",

}

Ruxanda, MM & Jensen, CS 2006, Efficient Similarity Retrieval in Music Databases. in Proceedings of the 13th International Conference on Management of Data. pp. 56-67, International Conference on Management of Data, New Delhi, India, 14/12/2006.

Efficient Similarity Retrieval in Music Databases. / Ruxanda, Maria Magdalena; Jensen, Christian Søndergaard.

Proceedings of the 13th International Conference on Management of Data. 2006. p. 56-67.

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

TY - GEN

T1 - Efficient Similarity Retrieval in Music Databases

AU - Ruxanda, Maria Magdalena

AU - Jensen, Christian Søndergaard

PY - 2006

Y1 - 2006

N2 - Audio music is increasingly becoming available in digital form, and the digital music collections of individuals continue to grow. Addressing the need for effective means of retrieving music from such collections, this paper proposes new techniques for content-based similarity search. Each music object is modeled as a time sequence of high-dimensional feature vectors, and dynamic time warping (DTW) is used as the similarity measure. To accomplish this, the paper extends techniques for time-series-length reduction and lower bounding of DTW distance to the multi-dimensional case. Further, the Vector Approximation file is adapted to the indexing of time sequences and to use a lower bound on the DTW distance. Using these techniques, the paper exploits the lack of a ground truth for queries to efficiently compute query results that differ only slightly from results that may be more accurate, but also are much more expensive, to compute. In particular, the paper demonstrates that aggressive use of time-series length reduction together with query expansion results in significant performance improvements while providing good, approximative query results.

AB - Audio music is increasingly becoming available in digital form, and the digital music collections of individuals continue to grow. Addressing the need for effective means of retrieving music from such collections, this paper proposes new techniques for content-based similarity search. Each music object is modeled as a time sequence of high-dimensional feature vectors, and dynamic time warping (DTW) is used as the similarity measure. To accomplish this, the paper extends techniques for time-series-length reduction and lower bounding of DTW distance to the multi-dimensional case. Further, the Vector Approximation file is adapted to the indexing of time sequences and to use a lower bound on the DTW distance. Using these techniques, the paper exploits the lack of a ground truth for queries to efficiently compute query results that differ only slightly from results that may be more accurate, but also are much more expensive, to compute. In particular, the paper demonstrates that aggressive use of time-series length reduction together with query expansion results in significant performance improvements while providing good, approximative query results.

KW - music similarity retrieval, time series, indexing

M3 - Article in proceeding

SP - 56

EP - 67

BT - Proceedings of the 13th International Conference on Management of Data

ER -

Ruxanda MM, Jensen CS. Efficient Similarity Retrieval in Music Databases. In Proceedings of the 13th International Conference on Management of Data. 2006. p. 56-67