CumInCAD is a Cumulative Index about publications in Computer Aided Architectural Design
supported by the sibling associations ACADIA, CAADRIA, eCAADe, SIGraDi, ASCAAD and CAAD futures

PDF papers
References
id ga0214
authors De Felice, F,Abbattista, F. and Scagliola, F.
year 2002
title GenOrchestra: An Interactive Evolutionary Agent for Musical Composition
source International Conference on Generative Art
summary GenOrchestra is a project involving the Dipartimento d’Informatica and Conservatorio di Musica “N. Piccinni” in Bari. This project concern a Creative Evolutionary System, based on Evolutionary Computation (EC) techniques, applied to the field of western tonal music. With GenOrchestra a novel way to evaluate the produced tunes is presented: indeed we adopt a hybrid solution composed for two kinds of fitness functions. The first, called technique fitness, evaluates the consonance degree between melodic, harmonic and rhythmic sections, moreover, it defines how well the rhythmic paths is organized into a coherent musical event. The second fitness function called human fitness, determine how well the tunes are perceived from a human audience, like in a concert. This task is accomplished by presenting the tunes on the Internet and then gathering the surfers evaluations in a database from which the system take the final population scoring. This, coupled with a no limited musical primordial soup, makes GenOrchestra a promising eclectic artificial composer. The ultimate goal of this project, currently in progress, is the development of a very human-like composer, which can produce music in any musical genre, and which is able to show a “personal style”. Samples will be soon available at http://valis.di.uniba.it/GenOrchestra/samples.html
series other
email febo74@libero.it
more http://www.generativeart.com/
full text file.pdf (353,720 bytes)
references Content-type: text/plain
last changed 2003/08/07 15:25
pick and add to favorite papersHOMELOGIN (you are user _anon_724671 from group guest) CUMINCAD Papers Powered by SciX Open Publishing Services 1.002