Presenter(s): Joseph Yaconelli − Math And Computer Science
Faculty Mentor(s): Robert Keller
Oral Session 3C
Research Area: Computer Science
Funding: National Science Foundation (NSF) Research Experience for Undergraduates (REU)
Building on previous work by Keller et al. in computer generated jazz solos using probabilistic grammars, this paper describes research extending the capabilities of the current learning process and grammar representation used in the Impro-Visor software with the concepts of motifs and motif patterns. An approach has been developed using clustering, best match search techniques, and probabilistic grammar rules to identify motifs and incorporate them into computer generated solos. The abilities of this technique are further expanded through the use of motif patterns. Motif patterns enable the learning of multiple lengths of motifs at once and induce coherence in generated solos by learning the patterns in which motifs
were used in a given set of solos. This approach is implemented as a feature of the Impro-Visor educational music software. Research has been done in other forms of pattern recognition and motif detection. However, this application of musical motif learning is a special case that requires vastly different techniques to accomplish due to music’s temporal nature, the variability of motifs both in length and melody, and the relatively short lifetime of motifs.