Lincs Lab extends its sincere congratulations to Cyprien Quéméneur, a recent alumnus, on successfully earning his Master of Applied Science (MASc) degree from Polytechnique Montréal. His thesis, “Federated Learning Methods for Real-Time Object Detection in Internet of Vehicles,” was accepted without revisions by the jury in late August. The examining committee included Professor Alejandro Quintero (President), Dr. Omar Abdul Wahab, and his supervisor, Professor Soumaya Cherkaoui.

Cyprien’s research delved into the potential of federated learning – a decentralized and communication-efficient approach designed to safeguard data privacy – for training advanced object detection models in connected and autonomous vehicles.

This innovative work culminated in the development of the FedPylot framework, an open-source lightweight MPI-based program designed to explore and simplify the federated training of real-time object detection models, particularly for the purpose of autonomous driving applications.