Publication: Sistema inteligente de composición armónica
Loading...
Identifiers
Publication date
2012-10
Defense date
2012-10-25
Authors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
Toda composición musical está sujeta a reglas que definen su estilo, regulando el esqueleto de la composición, del mismo modo que la experiencia musical adquirida durante años por el músico afectará a los patrones musicales de la obra, haciéndola diferente del resto de composiciones. Este trabajo pretende dar un nuevo enfoque a cómo afrontar la composición armónica automática aunando estos dos conceptos, reglas y experiencia, generando una estructura armónica, a partir de una línea melódica, por medio de un Sistema Experto y modificando cada una de las voces generadas en base a la experiencia, almacenada en forma de partituras, usando un sistema de Aprendizaje Automático basado en Instancias. El resultado será una melodía escrita en MusicXML que se aproxime al resultado obtenido por una persona en términos de calidad de expresión armónica. Se pretende demostrar que la composición de obras musicales expresivas, con estilo propio, no son exclusivas de las personas sino que son reproducibles utilizando técnicas de Inteligencia Artificial permitiendo que cualquier persona, sin necesidad de disponer de conocimientos musicales, pueda disfrutar de sus propias composiciones armónicas sin esfuerzo. _______________________________________________________________________________________________________________________
Any musical composition is subject to rules that define his style, regulating the composition framework, in the same way that the musical experience acquired over the years by the musician will affect musical patterns of the work, making it different from other compositions. This work tries to provide a new approach about how to address the automatic harmonic composition joining this two concepts, rules and experience, creating an harmonic framework, from a melodic line, through an Expert System and modifying each of the generated voices based on experience, stored in the form of scores, using a Machine Instance Based Learning. The result will be a melody written in MusicXML that comes close to the result obtained by a person in terms of harmonic expression quality. This works aims to show that the composition of expressive musical works, with a own style, doesn’t be exclusive of people but they are reproducible using Artificial Intelligence techniques allowing everyone, without need of music knowledge, to enjoy their own harmonic compositions effortlessly.
Any musical composition is subject to rules that define his style, regulating the composition framework, in the same way that the musical experience acquired over the years by the musician will affect musical patterns of the work, making it different from other compositions. This work tries to provide a new approach about how to address the automatic harmonic composition joining this two concepts, rules and experience, creating an harmonic framework, from a melodic line, through an Expert System and modifying each of the generated voices based on experience, stored in the form of scores, using a Machine Instance Based Learning. The result will be a melody written in MusicXML that comes close to the result obtained by a person in terms of harmonic expression quality. This works aims to show that the composition of expressive musical works, with a own style, doesn’t be exclusive of people but they are reproducible using Artificial Intelligence techniques allowing everyone, without need of music knowledge, to enjoy their own harmonic compositions effortlessly.
Description
Keywords
Inteligencia artificial, Sistemas expertos, Aprendizaje automático, Composiciones musicales