My work has been focused on the real-time control of additive synthesis under the direction of David Wessel, Director of CNMAT. I use neural networks, memory-based models, and related machine learning techniques, to model the spectral behaviors of musical instruments. Up to date, I have obtained very promising results with wind and string instruments as well as with the singing voice. I have presented this work at the 1998 International Computer Music Conference . 

Demos that illustrate the potential of this type of techniques and show what they already allow to do, are available online. 

• Removing the Time Axis from Spectral Model Analysis-Based Additive Synthesis: Neural Networks versus Memory-Based Machine Learning  in ICMC 1998 proceedings.
• " Motion of a pair of rigid ferrofluid drops in a rotating magnetic field" in Progr Colloid Polym Sci (1995) 98:124-12.

• My Resume is online.