Influence of low-level features extracted from rhythmic and harmonic sections on music genre classification

Aldona Rosner; Felix Weninger; Björn Schuller; Marcin Michalak; Bozena Kostek

doi:10.1007/978-3-319-02309-0_51

Influence of low-level features extracted from rhythmic and harmonic sections on music genre classification

Aldona Rosner, Felix Weninger, Björn Schuller, Marcin Michalak, Bozena Kostek

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

6 Scopus citations

Abstract

We present a comprehensive evaluation of the influence of “harmonic” and rhythmic sections contained in an audio file on automatic music genre classification. The study is performed using the ISMIS database composed of music files, which are represented by vectors of acoustic parameters describing low-level music features. Non-negative Matrix Factorization serves for blind separation of instrument components. Rhythmic components are identified and separated from the rest of the audio signals. Using such separated streams, it is possible to obtain information on the influence of rhythmic and harmonic components on music genre recognition. Further, the “original” audio feature vectors stemming from the non-separated signal are extended with such that base exclusively on drum and harmonic sections. The impact of these new parameters on music genre classification is investigated comparing the “basic” k-Nearest Neighbor classifier and Support Vector Machines.

Original language	English
Title of host publication	Man-Machine Interactions 3
Editors	Aleksandra Gruca, Tadeusz Czachórski, Stanisław Kozielski, Tadeusz Czachórski
Publisher	Springer Verlag
Pages	467-473
Number of pages	7
ISBN (Electronic)	9783319023083
DOIs	https://doi.org/10.1007/978-3-319-02309-0_51
State	Published - 2014
Externally published	Yes
Event	3rd International Conference on Man-Machine Interactions, ICMMI 2013 - Brenna, Poland Duration: 22 Oct 2013 → 25 Oct 2013

Publication series

Name	Advances in Intelligent Systems and Computing
Volume	242
ISSN (Print)	2194-5357

Conference

Conference	3rd International Conference on Man-Machine Interactions, ICMMI 2013
Country/Territory	Poland
City	Brenna
Period	22/10/13 → 25/10/13

Keywords

Drums separation
Instrument separation
MIR (Music Information Retrieval)
Music genre classification

Access to Document

10.1007/978-3-319-02309-0_51

Cite this

Rosner, A., Weninger, F., Schuller, B., Michalak, M., & Kostek, B. (2014). Influence of low-level features extracted from rhythmic and harmonic sections on music genre classification. In A. Gruca, T. Czachórski, S. Kozielski, & T. Czachórski (Eds.), Man-Machine Interactions 3 (pp. 467-473). (Advances in Intelligent Systems and Computing; Vol. 242). Springer Verlag. https://doi.org/10.1007/978-3-319-02309-0_51

Rosner, Aldona ; Weninger, Felix ; Schuller, Björn et al. / Influence of low-level features extracted from rhythmic and harmonic sections on music genre classification. Man-Machine Interactions 3. editor / Aleksandra Gruca ; Tadeusz Czachórski ; Stanisław Kozielski ; Tadeusz Czachórski. Springer Verlag, 2014. pp. 467-473 (Advances in Intelligent Systems and Computing).

@inproceedings{fed4772a4e1c4e4e810172ee81a7fd74,

title = "Influence of low-level features extracted from rhythmic and harmonic sections on music genre classification",

abstract = "We present a comprehensive evaluation of the influence of “harmonic” and rhythmic sections contained in an audio file on automatic music genre classification. The study is performed using the ISMIS database composed of music files, which are represented by vectors of acoustic parameters describing low-level music features. Non-negative Matrix Factorization serves for blind separation of instrument components. Rhythmic components are identified and separated from the rest of the audio signals. Using such separated streams, it is possible to obtain information on the influence of rhythmic and harmonic components on music genre recognition. Further, the “original” audio feature vectors stemming from the non-separated signal are extended with such that base exclusively on drum and harmonic sections. The impact of these new parameters on music genre classification is investigated comparing the “basic” k-Nearest Neighbor classifier and Support Vector Machines.",

keywords = "Drums separation, Instrument separation, MIR (Music Information Retrieval), Music genre classification",

author = "Aldona Rosner and Felix Weninger and Bj{\"o}rn Schuller and Marcin Michalak and Bozena Kostek",

note = "Publisher Copyright: {\textcopyright} Springer India 2014.; 3rd International Conference on Man-Machine Interactions, ICMMI 2013 ; Conference date: 22-10-2013 Through 25-10-2013",

year = "2014",

doi = "10.1007/978-3-319-02309-0_51",

language = "English",

series = "Advances in Intelligent Systems and Computing",

publisher = "Springer Verlag",

pages = "467--473",

editor = "Aleksandra Gruca and Tadeusz Czach{\'o}rski and Stanis{\l}aw Kozielski and Tadeusz Czach{\'o}rski",

booktitle = "Man-Machine Interactions 3",

}

Rosner, A, Weninger, F, Schuller, B, Michalak, M & Kostek, B 2014, Influence of low-level features extracted from rhythmic and harmonic sections on music genre classification. in A Gruca, T Czachórski, S Kozielski & T Czachórski (eds), Man-Machine Interactions 3. Advances in Intelligent Systems and Computing, vol. 242, Springer Verlag, pp. 467-473, 3rd International Conference on Man-Machine Interactions, ICMMI 2013, Brenna, Poland, 22/10/13. https://doi.org/10.1007/978-3-319-02309-0_51

Influence of low-level features extracted from rhythmic and harmonic sections on music genre classification. / Rosner, Aldona; Weninger, Felix; Schuller, Björn et al.
Man-Machine Interactions 3. ed. / Aleksandra Gruca; Tadeusz Czachórski; Stanisław Kozielski; Tadeusz Czachórski. Springer Verlag, 2014. p. 467-473 (Advances in Intelligent Systems and Computing; Vol. 242).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Influence of low-level features extracted from rhythmic and harmonic sections on music genre classification

AU - Rosner, Aldona

AU - Weninger, Felix

AU - Schuller, Björn

AU - Michalak, Marcin

AU - Kostek, Bozena

N1 - Publisher Copyright: © Springer India 2014.

PY - 2014

Y1 - 2014

N2 - We present a comprehensive evaluation of the influence of “harmonic” and rhythmic sections contained in an audio file on automatic music genre classification. The study is performed using the ISMIS database composed of music files, which are represented by vectors of acoustic parameters describing low-level music features. Non-negative Matrix Factorization serves for blind separation of instrument components. Rhythmic components are identified and separated from the rest of the audio signals. Using such separated streams, it is possible to obtain information on the influence of rhythmic and harmonic components on music genre recognition. Further, the “original” audio feature vectors stemming from the non-separated signal are extended with such that base exclusively on drum and harmonic sections. The impact of these new parameters on music genre classification is investigated comparing the “basic” k-Nearest Neighbor classifier and Support Vector Machines.

AB - We present a comprehensive evaluation of the influence of “harmonic” and rhythmic sections contained in an audio file on automatic music genre classification. The study is performed using the ISMIS database composed of music files, which are represented by vectors of acoustic parameters describing low-level music features. Non-negative Matrix Factorization serves for blind separation of instrument components. Rhythmic components are identified and separated from the rest of the audio signals. Using such separated streams, it is possible to obtain information on the influence of rhythmic and harmonic components on music genre recognition. Further, the “original” audio feature vectors stemming from the non-separated signal are extended with such that base exclusively on drum and harmonic sections. The impact of these new parameters on music genre classification is investigated comparing the “basic” k-Nearest Neighbor classifier and Support Vector Machines.

KW - Drums separation

KW - Instrument separation

KW - MIR (Music Information Retrieval)

KW - Music genre classification

UR - http://www.scopus.com/inward/record.url?scp=84927608036&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-02309-0_51

DO - 10.1007/978-3-319-02309-0_51

M3 - Conference contribution

AN - SCOPUS:84927608036

T3 - Advances in Intelligent Systems and Computing

SP - 467

EP - 473

BT - Man-Machine Interactions 3

A2 - Gruca, Aleksandra

A2 - Czachórski, Tadeusz

A2 - Kozielski, Stanisław

A2 - Czachórski, Tadeusz

PB - Springer Verlag

T2 - 3rd International Conference on Man-Machine Interactions, ICMMI 2013

Y2 - 22 October 2013 through 25 October 2013

ER -

Rosner A, Weninger F, Schuller B, Michalak M, Kostek B. Influence of low-level features extracted from rhythmic and harmonic sections on music genre classification. In Gruca A, Czachórski T, Kozielski S, Czachórski T, editors, Man-Machine Interactions 3. Springer Verlag. 2014. p. 467-473. (Advances in Intelligent Systems and Computing). doi: 10.1007/978-3-319-02309-0_51

Influence of low-level features extracted from rhythmic and harmonic sections on music genre classification

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this