VOISE: Learning to Segregate Voices in Explicit and Implicit Polyphony

Finding multiple occurrences of themes and patterns in music can be hampered due to polyphonic textures. This is caused by the complexity of music that weaves multiple independent lines of music together. We present and demonstrate a system, VoiSe, that is capable of isolating individual voices in both explicit and implicit polyphonic music. VoiSe is designed to work on a symbolic representation of a music score, and consists of two components: a same-voice predicate implemented as a learned decision tree, and a hard-coded voice numbering algorithm.

• Type: Conference
• Year: 2005
• Where: ISMIR
• Authors: Phillip B. Kirlin, Paul E. Utgoff
• Keywords: Implicit Polyphonic Music, Information Retrieval, ISMIR 2005, Multiple Independent Lines, Polyphonic Textures