Senior Researcher
CurrentAs the most experienced ML scientist in the lab, I take part in several multidisciplinary teams to apply AI to media, telecommunication networks or robotics. Within a small team, I also drove the development of new deep learning architectures for embedded devices. I also supervise postdocs, interns and PhD students, and co-lead our contribution to the EU-funded AI4media Center of Excellence.Some achievements:- Estimating network performance with GNN, decreasing error from over 100% to 1.63%- Recurrent architecture that adapts to available computation/energy, can reduce computation by 85% without affecting accuracy