François Pachet: Artificial Creativity with Markov Constraints
In the context of the Flow Machine project, I introduce the problem of style imitation with user constraints. Such a problem arises naturally as soon as one wants to enable users to generate sequential content (music, text, drawings) in the style of an author, that also satisfy their own constraints (ideas, patterns). Statistical models based on Markov chains are in principle well adapted to that aim but the traditional generation techniques (random walk) cannot cope with additional, non Markovian constraints, by definition. I will describe the approach we have been following to solve this issue and various recent results we obtained, notably to produce Markov chains with global constraints (such as meter or maxOrder, related to plagiarism / innovation). I illustrate these ideas with several music and text generation applications. Francois Pachet - Director of Sony Computer Science Laboratory Paris Francois Pachet received his Ph.D. and Habilitation degrees from Paris 6 University (UPMC). He is a Civil Engineer (Ecole des Ponts and Chaussées) and was Assistant Professor in Artificial Intelligence and Computer Science, at Paris 6 University, until 1997. He is now director of SONY Computer Science Laboratory Paris, where he leads the music research team. The music team conducts research on interactive music listening, composition and performance. Since its creation, the team developed several award winning technologies (constraint-based spatialization, intelligent music scheduling using metadata) and systems (MusicSpace, PathBuilder, Continuator for interactive music improvisation, etc.). His current goal is to create a new generation of authoring tools able to boost individual creativity. These tools, called Flow Machines, abstract "style" from concrete corpora (text, music, etc.), and turn it into a malleable substance that acts as a texture. Applications range from music composition to text or drawing generation and probably much more. The A-Talks are gathering world-reknowned experts in fields related to robotics & AI, presenting ground breaking ideas in various domain from robotics and computer science, to psychology, language or social sciences. We want to bring to the audience fresh new ideas, just beyond the level of popular science into the real core of today's top most advanced research in the field. Our ambition is that if you listen to an A-Talk, there is a good chance you'll hear ideas that you've never heard before!