Boundary Attack++: Query-Efficient Decision-Based Adversarial Attack

41 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Adversarial Training and Robustness for Multiple Perturbations (2019, methodology): “, PGD [25] and our SLIDE attack with 100 steps and 40 restarts (20 restarts on CIFAR10), as well as Carlini and Wagner’s l2-attack [6] (C&W), and an l1-variant—EAD [8]; (2) to detect gradient-masking, we use decision-based attacks: the Boundary Attack [3] for l2, the Pointwise Attack [31] for l1, an”
Stateful Detection of Black-Box Adversarial Attacks (2019, background): “Since its publication, there have been multiple proposals to improve this attack [24, 22].”
Adversarial Robustness Toolbox v1.0.0 (2018, background): “…[10], Section 6.16 • Decision Tree Attack [45], Section 6.17 • High Conﬁdence Low Uncertainty (HCLU) Attack [24], Section 6.18 • HopSkipJump Attack [15], Section 6.19 ART has an abstract class Attack in the art.attacks.attack module which implements a common interface for any of the particular…”