The present study investigates the modeling of the perception of time signature and phase using grammar-based and memorybased approaches. It explores how far note-length can be used as the sole input to a model of perception of heard music. Two models are developed: one uses a rule-based grammar and the other uses a combination of a rule-based grammar and a memory-based approach. The relative success of the two models is compared and evaluated. This study explores one dialect, solo string music by Bach: The unaccompanied Suites for Violoncello, BWV 1007-1012 and unaccompanied Partitas and Sonatas for Violin BWV 1001-1006. It is shown that the results of two approaches improve over previous models such as those of Longuet-Higgins and Steedman (1971) and Temperley and Sleator (1999) and that the combination of the two approaches is most successful.
CITATION STYLE
Spiro, N. (2002). Combining grammar-based and memory-based models of perception of time signature and phase. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2445, pp. 183–194). Springer Verlag. https://doi.org/10.1007/3-540-45722-4_17
Mendeley helps you to discover research relevant for your work.