Abstract
In this paper we introduce multi-label ferns, and apply this technique for automatic classification of musical instruments in audio recordings. We compare the performance of our proposed method to a set of binary random ferns, using jazz recordings as input data. Our main result is obtaining much faster classification and higher F-score. We also achieve substantial reduction of the model size. © 2014 Springer International Publishing.
Cite
CITATION STYLE
Kursa, M. B., & Wieczorkowska, A. A. (2014). Multi-label ferns for efficient recognition of musical instruments in recordings. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8502 LNAI, pp. 214–223). Springer Verlag. https://doi.org/10.1007/978-3-319-08326-1_22
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
