Communication dans un congrès, AO, GALaC, Informatique

Provably Safeguarding a Classifier from OOD and Adversarial Samples: an Extreme Value Theory Approach

Nicolas Atienza, Christophe Labreuche, Johanne Cohen, Michèle Sebag. Provably Safeguarding a Classifier from OOD and Adversarial Samples: an Extreme Value Theory Approach. ICLR 2025 - The Thirteenth International Conference on Learning Representations, Apr 2025, Singapore (SG), Singapore. ⟨hal-04922382⟩

Publié le