In this paper, we use Hidden Markov Models (HMM) andMel-Frequency Cepstral Coefficients (MFCC) to build statistical modelsof classical music composers directly from the music datasets. Severalmusical pieces are divided by instruments (String, Piano, Chorus, Or-chestra), and, for each instrument, statistical models of the composersare computed.We selected 19 dierent composers spanning four centuriesby using a total number of 400 musical pieces. Each musical piece is classied as belonging to a composer if the corresponding HMM gives thehighest likelihood for that piece. We show that the so-developed modelscan be used to obtain useful information on the correlation between thecomposers. Moreover, by using the maximum likelihood approach, wealso classified the instrumentation used by the same composer. Besidesas an analysis tool, the described approach has been used as a classifier. This overall originates an HMM-based framework for supportingaccurate classification of music datasets. On a dataset of String Quartetmovements, we obtained an average composer classification accuracy ofmore than 96%. As regards instrumentation classification, we obtained anaverage classification of slightly less than 100% for Piano, Orchestra andString Quartet. In this paper, the most significant results coming fromour experimental assessment and analysis are reported and discussed indetail.

An HMM-Based Framework for Supporting Accurate Classification of Music Datasets

CUZZOCREA, Alfredo Massimiliano
;
2017

Abstract

In this paper, we use Hidden Markov Models (HMM) andMel-Frequency Cepstral Coefficients (MFCC) to build statistical modelsof classical music composers directly from the music datasets. Severalmusical pieces are divided by instruments (String, Piano, Chorus, Or-chestra), and, for each instrument, statistical models of the composersare computed.We selected 19 dierent composers spanning four centuriesby using a total number of 400 musical pieces. Each musical piece is classied as belonging to a composer if the corresponding HMM gives thehighest likelihood for that piece. We show that the so-developed modelscan be used to obtain useful information on the correlation between thecomposers. Moreover, by using the maximum likelihood approach, wealso classified the instrumentation used by the same composer. Besidesas an analysis tool, the described approach has been used as a classifier. This overall originates an HMM-based framework for supportingaccurate classification of music datasets. On a dataset of String Quartetmovements, we obtained an average composer classification accuracy ofmore than 96%. As regards instrumentation classification, we obtained anaverage classification of slightly less than 100% for Piano, Orchestra andString Quartet. In this paper, the most significant results coming fromour experimental assessment and analysis are reported and discussed indetail.
Artificial Intelligence
Classification
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/20.500.11770/312800
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact