logo Idiap Research Institute        
 [BibTeX] [Marc21]
Predictive Models for Music
Type of publication: Idiap-RR
Citation: paiement:rr08-51
Number: Idiap-RR-51-2008
Year: 2008
Institution: IDIAP
Note: Submitted for publication
Abstract: Modeling long-term dependencies in time series has proved very difficult to achieve with traditional machine learning methods. This problem occurs when considering music data. In this paper, we introduce generative models for melodies. We decompose melodic modeling into two subtasks. We first propose a rhythm model based on the distributions of distances between subsequences. Then, we define a generative model for melodies given chords and rhythms based on modeling sequences of Narmour features. The rhythm model consistently outperforms a standard Hidden Markov Model in terms of conditional prediction accuracy on two different music databases. Using a similar evaluation procedure, the proposed melodic model consistently outperforms an Input/Output Hidden Markov Model. Furthermore, sampling these models given appropriate musical contexts generates realistic melodies.
Userfields: ipdmembership={learning},
Projects Idiap
Authors Paiement, Jean-Fran├žois
Grandvalet, Yves
Bengio, Samy
Added by: [UNK]
Total mark: 0
  • paiement-idiap-rr-08-51.pdf
  • paiement-idiap-rr-08-51.ps.gz