formal verification

Caractériser des propriétés de confiance d’IA avec Why3

Nous proposons de faire la démonstration de CAISAR, un logiciel libre en développement au sein du laboratoire de sûreté et sécurité des logiciels du CEA LIST. CAISAR est une plateforme permettant de caractériser la sûreté des logiciels résultant d’un …

IDESSAI 2021

Formal verification of deep neural networks: theory and practice; a tutorial I gave during the joint INRIA-DFKI 2021 summer school.

PFIA 2020

Theory and practice of deep learning verification; a tutorial I gave during the PFIA 2020 conference

CAMUS: A Framework to Build Formal Specifications for Deep Perception Systems Using Simulators

The topic of provable deep neural network robustness has raised considerable interest in recent years. Most research has focused on adversarial robustness, which studies the robustness of perceptive models in the neighbourhood of particular samples. …