Saturday 23 December 2006, by jmc
All the versions of this article:
Alain Bonardi is a senior lecturer at Paris 8 University and belongs as a researcher to the Maison des Sciences de l’Homme Paris-Nord. His research work mainly concerns the application of artificial intelligence to performing arts (opera and theatre). For the production of the theatre play La traversée de la nuit by Geneviève de Gaulle, he developed a system based on neural networks to learn the comedian’s voice and generate images through a multi-agent system. For the interactive open form opera Alma Sola (creation of the full version in october 2005 in Le Cube), he created an opera continuation system based on Hidden Markov Models.
He mainly developed two research approaches linked to visualisation of sound:
Temporal decorrelation of sound-image mapping. The notion of memory, associated to various temporal scales, is a keypoint of this approach. For instance, in La traversée de la nuit, emotional states acknowledged in the comedian’s voice are taken into account in the « sensitivity » model of the agents which deal with image creation.
Data representation : we are working on a virtual assistant project for performers and stage directors. It will enable to classify intention expressions from video and sound captations of a performer, and to build a temporal synoptic scheme (see Isis Truck)
A. Bonardi, New Approaches of Theatre and Opera Directly Inspired by Interactive Data-mining, Sound & Music Computing Conference (SMC’04), pages 1-4, Paris, 20-22 octobre 2004.