(1) PAPER TITLES the complete title of one (or more) paper(s) published in the open literature describing the work that the author claims describes a human-competitive result. [1] Exploiting Functional Relationships in Musical Composition [2] Scaffolding for Interactively Evolving Novel Drum Tracks for Existing Songs ----------------------------------------------------------------------- -------------------- (2) AUTHORS the name, complete physical mailing address, e-mail address, and phone number of EACH author of EACH paper, Amy K. Hoover School of Electrical Engineering and Computer Science University of Central Florida 4000 Central Florida Blvd. Orlando, FL 32816-2362 ahoover@eecs.ucf.edu 727-460-9764 Michael P. Rosario School of Electrical Engineering and Computer Science University of Central Florida 4000 Central Florida Blvd. Orlando, FL 32816-2362 michael.rosario@yahoo.com 407-823-4289 Kenneth O. Stanley School of Electrical Engineering and Computer Science University of Central Florida 4000 Central Florida Blvd. Orlando, FL 32816-2362 kstanley@eecs.ucf.edu 407-473-0072 ---------------------------------------------------------------------------- --------------- (3) CORRESPONDING AUTHOR the name of the corresponding author (i.e., the author to whom notices will be sent concerning the competition). Amy K. Hoover ------------------------------------------------------------------------------------------- (4) ABSTRACTS the abstract of the paper(s). [1] The ability of gifted composers such as Mozart to create complex multipart musical compositions with relative ease suggests a highly efficient mechanism for generating multiple parts simultaneously. Computational models of human music composition can potentially shed light on how such rapid creativity is possible. This paper proposes such a model based on the idea that the multiple threads of a song are temporal patterns that are functionally related, which means that one instrument's sequence is a function of another's. This idea is implemented in a program called NEAT Drummer that interactively evolves a type of artificial neural network (ANN) called a Compositional Pattern Producing Network (CPPN), which represents the functional relationship between the instruments and drums. The main result is that richly textured drum tracks that tightly follow the structure of the original song are easily generated because of their functional relationship to it. [2] A major challenge in computer-generated music is to produce music that sounds natural. This paper introduces NEAT Drummer, which takes steps toward natural creativity. NEAT Drummer evolves a kind of artificial neural network called a Compositional Pattern Producing Network (CPPN) with the NeuroEvolution of Augmenting Topologies (NEAT) method to produce drum patterns. An important motivation for this work is that instrument tracks can be generated as a function of other song parts, which, if written by humans, thereby provide a scaffold for the remaining auto-generated parts. Thus, NEAT Drummer is initialized with inputs from an existing MIDI song and through interactive evolution allows the user to evolve increasingly appealing rhythms for that song. This paper explains how NEAT Drummer processes MIDI inputs and outputs drum patterns. The net effect is that a compelling drum track can be automatically generated and evolved for any song. ------------------------------------------------------------------------------------------- (5) CRITERIA a list containing one or more of the eight letters (A, B, C, D, E, F, G, or H) that correspond to the criteria (see above) that the author claims that the work satisfies. F, G ------------------------------------------------------------------------------------------- (6) STATEMENT ON WHY THESE CRITERIA ARE SATISFIED a statement stating why the result satisfies the criteria that the contestant claims (see examples of statements of human-competitiveness as a guide to aid in constructing this part of the submission). Automated music generation is among the holy grails of computational creativity. The prospect of computers composing or helping to compose music that genuinely rivals that of professional human artists could revolutionize the music and entertainment industries and radically change the way musical entertainment is generated. It is widely acknowledged that a major problem is automated music generation is that while it can produce locally convincing motifs, it generally fails to capture the global structure that makes music interesting (McCormack, 2005; Husbands et al., 2007; full references provided later). NEAT Drummer is a program that automatically generates natural sounding percussion tracks for existing human compositions that actually respects and understands global structure. Based on the existing song parts (i.e. the melody, harmony, and bass), NEAT Drummer generates a drum pattern population from which the user can choose to further evolve. Each drum pattern flows with the individual contours of the song. That is, as the global structure transitions, so does the drum pattern transition seamlessly. Thus NEAT Drummer is a milestone in automated music generation, introducing an entirely new theoretical framework for generating convincing global structure for new tracks within an existing musical composition. The submitted results are drum patterns generated by NEAT Drummer specifically for the folk songs Johnny Cope and Oh! Susanna. The songs can be heard at: http://eplex.cs.ucf.edu/neatmusic/JohnnyCopeG1.mp3 http://eplex.cs.ucf.edu/neatmusic/JohnnyCopeG11.mp3 http://eplex.cs.ucf.edu/neatmusic/Oh!SusannaG25.mp3 Attention should be paid in particular to the drums, which continually change and improvise as the song transitions, yet sound natural and respectful of it structure both locally and globally. These drum pattern accompaniments satisfy both criteria F and G: F: “The result is equal to or better than a result that was considered an achievement in its field at the time it was first discovered.” The demo songs above were chosen intentionally because they were originally sequenced by respected folk musician Barry Taylor without any drum tracks. Barry Taylor also specifically granted us permission to include his drum-less music in our experiments. Some examples of Barry Taylor’s work are at these sites: http://www.contemplator.com/tunebook/index.htm http://members.shaw.ca/tunebook/ To support criterion F, we quote directly from Barry Taylor himself on the results from NEAT Drummer on his own songs: “Those samples were amazing! … I have heard some ‘auto drum’ software in the past and the result, at least when applied to old traditional tunes, is usually appalling. I must confess that I was a little wary of listening to the results of your application applied to *my* sequences, but your work gives me renewed hope that great quality can indeed be achieved! Keep up the superb work! I wish you continued success in your project...and I'm thrilled to contribute in this small way!” Note that Mr. Taylor cites prior “auto drum” software for comparison, demonstrating that NEAT Drummer surpasses prior results in the field, according to an experienced musician. Furthermore, you can listen to the provided tracks to judge for yourself. G: “The result solves a problem of indisputable difficulty in its field.” As noted above, the problem of capturing global structure in automated music is widely acknowledged. Respected researchers such as in the publications listed below (which were cited above) have highlighted this problem of indisputable difficulty: -McCormack, J. (2005). Open problems in evolutionary music and art. In Proceedings of Applications of Evolutionary Computing, (EvoMUSART 2005), volume 3449 of Lecture Notes in Computer Science, pages 428{436, Berlin, Germany. Springer Verlag. -Husbands, P., Copley, P., Eldridge, A., and Mandelis, J. (2007). An introduction to evolutionary computing for musicians. In Miranda, E. R. and Biles, J. A., editors, Evolutionary Computer Music, chapter 1, pages 1{27. Springer-Verlag New York, Inc., Secaucus, NJ, USA Barry Taylor’s judgment in combination with these citations shows that the program solves a problem of indisputable difficulty. Furthermore, our 2008 EvoMUSART paper won the Best Paper Award at the symposium, demonstrating that the evolutionary music and art community agrees (i.e. there is no dispute) that NEAT Drummer solves an important problem. ---------------------------------------------------------- --------------------------------- (7) CITATIONS a full citation of the paper (that is, author names; publication date; name of journal, conference, technical report, thesis, book, or book chapter; name of editors, if applicable, of the journal or edited book; publisher name; publisher city; page numbers, if applicable). [1] A. K. Hoover and K. O. Stanley. Exploiting functional relationships in musical composition. Connection Science Special Issue on Music, Brain, & Cognition, 21(2):227-251, accepted to appear June 2009. (exact form of final paper accepter before May 2008) [2] A. K. Hoover, M. P. Rosario, and K. O. Stanley. Scaffolding for interactively evolving novel drum tracks for existing songs. In M. G. et. al., editor, Proceedings of the Sixth European Workshop on Evolutionary and Biologically Inspired Music, Sound, Art and Design (EvoMUSART 2008), pages 412-422. Springer, March 2008. WINNER OF THE BEST PAPER AWARD IN EVOLUTIONARY MUSIC AND ART Please note: We include this second paper here even though it appeared a couple months before the time period of the competition because it describes an earlier version of NEAT Drummer. However, because the first paper (above) is within the time period of the competition and this paper is related to it, we felt it made sense to include it too for reference. Reference [1] is the first paper ever published on the full NEAT Drummer methodology, while this paper [2] is an earlier version that did not describe the entire methodology. If needed, this second reference [2] can be removed and the submission still stands with reference [1]. ---------------------------------------------------------------------------------------- --- (8) STATEMENT ON PRIZE MONEY a statement either that “any prize money, if any, is to be divided equally among the co-authors” OR a specific percentage breakdown as to how the prize money, if any, is to be divided among the co-authors. Any prize money, if any, is to be divided equally among the co-authors. ---------------------------------------------------------------------------------------- --- (9) STATEMENT ON WHY THIS ENTRY IS THE BEST a statement stating why the judges should consider the entry as “best” in comparison to other entries that may also be “human-competitive.” NEAT Drummer is a highly original idea in evolutionary music generation, showing how global structure can be extracted from existing scaffolding within a song, which is why it is endorsed by Barry Taylor and winner of a best paper award (which does imply that it is best in its category). HUMIES winners are traditionally contributions to science and engineering. Yet the arts are still profoundly dominated by human beings. Because of the universal appeal of music and art, to show that computers are ascendant within the arts can appeal to a broad cross-section of society that normally would not be aware of scientific advancements, thereby helping to promote the potential of evolutionary computation to compete with human accomplishment. If artificial technologies are really “human competitive” then they should be competitive across the spectrum of human achievement, not just within technical fields. It would raise suspicion if the claim is made that a particular technology competes with humans yet its accomplishments consistently exclude among the most human of all human endeavors, which is our creative side. Thus the advance represented by NEAT Drummer, which significantly impacts computer generated music, and has earned the recognition of both musicians and the evolutionary art and music community, is a milestone that can benefit the aims of the HUMIES and complement the significant achievements that evolutionary computation has lent to science and engineering.