Pierre-Yves LAGRAVE (THALES RESEARCH & TECHNOLOGY)
Frédéric BARBARESCO (THALES LAND & AIR SYSTEMS)
30th November 2021 - 4:00 pm - 5:00 pm (GST)
Title: | Introduction to Robust Machine Learning with Geometric Methods for Defense Applications |
Abstract: | This talk aims at motivating the use of geometrically informed Machine Learning algorithms for Defense applications by providing intuitions with respect to their underlying mechanisms and by shedding light on successful applications such as remote sensing imagery, radar Doppler signal processing, trajectory prediction, physical model simulation, and kinematics recognition. We in particular discuss the use of Equivariant Neural Networks (ENN), which achieve geometrical robustness by-design and which also appear more robust to adversarial attacks. We will also highlight how Lie Group Statistics and Machine Learning techniques can be used to process data in their native geometry. Both technologies have a wide range of applications for the Defense industry, and we generally believe that exploiting the data geometry and the underlying symmetries is key to the design of efficient, reliable, and robust AI-based Defense systems. |
Bio: | Frédéric BARBARESCO Senior THALES Expert in Artificial Intelligence at the Technical Department of THALES Land & Air Systems. SMART SENSORS Segment Leader for the THALES Corporate Technical Department (Key Technology Domain "Processing, Control & Cognition"). THALES representative at the AI Expert Group of ASD (AeroSpace and Defense Industries Association of Europe). 2014 Aymée Poirson Prize of the French Academy of Science for the application of science to industry. Ampère Medal, Emeritus Member of the SEE, and President of the SEE ISIC club "Information and Communication Systems Engineering". General Chair of the "Geometric Science of Information" international conference (https://franknielsen.github.io/GSI/) and SPRINGER Book Editor on “Geometric Structures of Statistical Physics, Information Geometry, and Learning” (https://link.springer.com/book/10.1007%2F978-3-030-77957-3) Pierre-Yves LAGRAVE Pierre-Yves Lagrave graduated from the French engineering school Ecole Nationale des Ponts et Chaussées in 2011, where he majored in Applied Mathematics and Computer Science. He started his career as a quantitative analyst in the financial industry (Société Générale, JP Morgan), where he was eventually leading a model risk team in London. He took a research engineer position at the French Ministry of Armed Forces in 2016 where he developed mathematical tools for software and cyber security matters. He then joined Thales Research and Technology as a research engineer in January 2020 and focuses on Trustworthy Artificial Intelligence since then. His areas of expertise and research interests include Safe and Secure Machine Learning, Geometric Deep Learning, Stochastic Modelling and Differential Geometry. |