In this paper, we use Hidden Markov Models (HMM) and Mel-Frequency Cepstral Coefficients (MFCC) to build statistical models of classical music composers directly from the music datasets. Several musical pieces are divided by instruments (String, Piano, Chorus, Or- chestra), and, for each instrument, statistical models of the composers are computed.We selected 19 dierent composers spanning four centuries by using a total number of 400 musical pieces. Each musical piece is classi ed as belonging to a composer if the corresponding HMM gives the highest likelihood for that piece. We show that the so-developed models can be used to obtain useful information on the correlation between the composers. Moreover, by using the maximum likelihood approach, we also classified the instrumentation used by the same composer. Besides as an analysis tool, the described approach has been used as a classifier. This overall originates an HMM-based framework for supporting accurate classification of music datasets. On a dataset of String Quartet movements, we obtained an average composer classification accuracy of more than 96%. As regards instrumentation classification, we obtained an average classification of slightly less than 100% for Piano, Orchestra and String Quartet. In this paper, the most significant results coming from our experimental assessment and analysis are reported and discussed in detail.
An HMM-Based Framework for Supporting Accurate Classification of Music Datasets
CUZZOCREA, Alfredo Massimiliano;MUMOLO, ENZO;
In corso di stampa
Abstract
In this paper, we use Hidden Markov Models (HMM) and Mel-Frequency Cepstral Coefficients (MFCC) to build statistical models of classical music composers directly from the music datasets. Several musical pieces are divided by instruments (String, Piano, Chorus, Or- chestra), and, for each instrument, statistical models of the composers are computed.We selected 19 dierent composers spanning four centuries by using a total number of 400 musical pieces. Each musical piece is classi ed as belonging to a composer if the corresponding HMM gives the highest likelihood for that piece. We show that the so-developed models can be used to obtain useful information on the correlation between the composers. Moreover, by using the maximum likelihood approach, we also classified the instrumentation used by the same composer. Besides as an analysis tool, the described approach has been used as a classifier. This overall originates an HMM-based framework for supporting accurate classification of music datasets. On a dataset of String Quartet movements, we obtained an average composer classification accuracy of more than 96%. As regards instrumentation classification, we obtained an average classification of slightly less than 100% for Piano, Orchestra and String Quartet. In this paper, the most significant results coming from our experimental assessment and analysis are reported and discussed in detail.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.