Related Experiment Videos

Unsupervised analysis of polyphonic music by sparse coding

Samer A Abdallah1, Mark D Plumbley

  • 1Department of Electronic Engineering, Queen Mary, University of London, London E1 4NS, UK. samer.abdallah@elec.qmul.ac.uk

Summary

This study introduces a novel data-driven probabilistic model for polyphonic music analysis. The system efficiently decomposes musical spectra to identify individual notes, learning directly from polyphonic music data.

Related Concept Videos