Thursday, 13th of December

11h11 (room R2014, 660 building) (see location)

Julien Girard

(TAU / CEA-list)

Title: A short introduction to formal methods and their applications for Robust Deep networks


Abstract

Deep neural networks present strong vulnerabilities against adversarial examples, model theft and privacy of datasets.
On the other side, software verification is a well-researched field, with about 40 years of cumulated expertise, tools and techniques.
We aim to present two techniques widely used in software verification: Abstract Interpretation and Satisfactory Modulo Theories solvers, and how we can leverage them to obtain provably more robust deep neural networks.



Contact: guillaume.charpiat at inria.fr
All TAU seminars: here