Probabilistic Jacobian-Based Saliency Maps Attacks

Théo Combey (1) , António Loison (1) , Maxime Faucher (1) , Hatem Hajri (2)

1 CentraleSupélec
2 IRT SystemX

This paper introduces simple, faster and more efficient versions of the known targeted and untargeted Jacobian-based Saliency Map Attacks (JSMA). Despite creating adversarial examples with a higher average L 0 distance than the state-of-the-art Carlini-Wagner attack, the new versions of JSMA have a significant speed advantage over this attack, making them very convenient for L 0 real-time robustness testing of neural network classifiers.

Mots clés: Jacobian-based Saliency Map; adversarial attacks; deep neural network classifiers; MNIST; CIFAR-10; GTSRB

Domaines: Sciences de l'ingénieur [physics]

Théo Combey, António Loison, Maxime Faucher, Hatem Hajri. Probabilistic Jacobian-Based Saliency Maps Attacks. Machine Learning and Knowledge Extraction, 2020, 2 (4), pp.558 - 578. ⟨10.3390/make2040030⟩. ⟨hal-03085884⟩