Michaël Perrot received the 2017 Best Phd Thesis award in Artificial Intelligence

Michaël Perrot received the 2017 Best Phd Thesis award in Artificial Intelligence from the French society for Artificial Intelligence (AFIA http://www.afia.asso.fr) which promotes the best national research work of the year in AI.

Michaël was a Phd student in the Laboratoire Hubert Curien where he has been supervised by Pr. Amaury Habrard; he defended his PhD in December 2016.

Michaël received this award for his contributions in machine learning and more precisely in metric learning, a subfield of representation learning. This topic concerns the problem of automatically learning distances or similarities from data that can be then used in classification, prediction or clustering systems. The Phd of Michaël has focused on the study of new frameworks and algorithms able to control the behavior of a metric learned in a wide range of contexts such as classification, transfer of metrics, and the incorporation of local or global constraints. He got many publications in top-tier international conferences, like NIPS and ICML, and addressed applications in computer vision such as perceptual color differences, domain adaptation or seamless copy in images.

Michaël is now a post-doctoral researcher at the Max-Planck Institute at Tübingen, Germany.