Canonical foliations of neural networks: application to robustness

no code implementations • 2 Mar 2022 • Eliot Tron, Nicolas Couellan, Stéphane Puechmorel

The result show that the proposed attack is more efficient at all levels of available budget for the attack (norm of the attack), confirming that the curvature of the transverse neural network FIM foliation plays an important role in the robustness of neural networks.

Adversarial Attack

Classifying histograms of medical data using information geometry of beta distributions

no code implementations • 3 Jun 2020 • Alice Le Brigant, Nicolas Guigui, Sana Rebbah, Stéphane Puechmorel

In this paper, we use tools of information geometry to compare, average and classify histograms.