"Representing Style: the contribution of Markov constraints" (François Pachet)
Wednesday May 2 — h 4:00 pm
Representing Style: the contribution of Markov Constraints
The motivation of this work is to generate sequences (typically music or text)
that imitate a given "style".
A traditional approach to represent style is to use statistical methods such as Markov processes.
Markov processes capture adequately short-term
dependencies in temporal sequences, but are very difficult to control.
Imposing arbitrary properties on Markov chains is indeed paradoxical as it often goes in the way of the basic Markov hypothesis of limited temporal dependency.
This talk introduces our work on Markov constraints: a new class of global constraints whose goal is to reformulate Markov processes in the framework of CSP, to enable users to generate sequences "in the style" of a given corpus,
while satisfying arbitrary control criteria.
We will give examples in music and text generation, and relate this work to the ERC funded "Flow Machines" project.
Introduction by Giovanni Emanuele Corazza
DEIS - Università di Bologna & MIC Marconi Institute for Creativity
Organized by Mirko Degli Esposti
Dipartimento di Matematica - Università di Bologna