What/when causal expectation modelling applied to percussive audio

TitleWhat/when causal expectation modelling applied to percussive audio
Publication TypeJournal Article
Year of Publication2009
AuthorsHazan, A., Brossier P., Marxer R., & Purwins P.
Journal TitleThe Journal of the Acoustical Society of America, vol. 123, issue 5, p. 3800
AbstractA causal system to represent a stream of music into musical events, and generate further expected events, is presented. Starting from an auditory front-end which extracts low and mid-level features, an unsupervised clustering process builds and maintains a set of symbols aimed at representing musical stream events using both timbre and time descriptions. These symbols are then processed by an expectation module using a multiscale technique based on N-Grams. To characterize the ability of the system to generate an expectation that matches both ground truth and system transcription, we introduce several measures that take into account the uncertainty associated with the unsupervised encoding of the musical sequence. We compare three approaches to combine timing (when) and timbre (what) expectation.
preprint/postprint documenthttp://hdl.handle.net/10230/34478
Final publicationhttp://doi.org/10.1080/09540090902733764