Automatic Recognition of Key Modulations in Symbolic Musical Pieces Using Information Theory

Autor:	Michele Della Ventura
Rok vydání:	2021
Předmět:	Computer science Movement (music) business.industry computer.software_genre Key (music) Music information retrieval Natural (music) Musical composition Polyphony Artificial intelligence business Tonality computer Period (music) Natural language processing
Zdroj:	Lecture Notes in Networks and Systems ISBN: 9783030821920 IntelliSys (1)
DOI:	10.1007/978-3-030-82193-7_56
Popis:	Extracting information of musical attributes such as harmony and key modulation from the symbolic music text is a critical process in Music Information Retrieval (MIR) systems. This research paper presents a new approach to automatically detect key modulations in a musical piece considered in its symbolic musical level. The method does not entail any preset rules respect to the musical grammar. The Markov process is used in order to identify the relations between the sounds and their movement within the harmonic structure. The information theory is used to define the tonality (or key) when sounds are changed by means of accidentals (sharp, flat or natural). The method has been tested on tonal polyphonic compositions and gave encouraging results, with an accuracy ranging between 82% and 91% on the base of the historical period of the analyzed music composition. Future improvements of the method are discussed briefly at the end of the paper.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::cc98fc92043124c03782d5a374f75b80 https://doi.org/10.1007/978-3-030-82193-7_56 Zobrazit plný text záznamu