OMG-ATTACK: Self-Supervised On-Manifold Generation of Transferable Evasion Attacks

O Bar Tal, A Haviv, AH Bermano - Proceedings of the IEEE …, 2023 - openaccess.thecvf.com
Evasion Attacks (EA) are used to test the robustness of trained neural networks by distorting
input data to misguide the model into incorrect classifications. Creating these attacks is a …

Imperceptible and sparse adversarial attacks via a dual-population-based constrained evolutionary algorithm

Y Tian, J Pan, S Yang, X Zhang… - IEEE transactions on …, 2022 - ieeexplore.ieee.org
The sparse adversarial attack has attracted increasing attention due to the merit of a low
attack cost via changing a small number of pixels. However, the generated adversarial …

An adaptive black-box defense against trojan attacks (trojdef)

G Liu, A Khreishah, F Sharadgah… - IEEE Transactions on …, 2022 - ieeexplore.ieee.org
Trojan backdoor is a poisoning attack against neural network (NN) classifiers in which
adversaries try to exploit the (highly desirable) model reuse property to implant Trojans into …

Explore Adversarial Attack via Black Box Variational Inference

C Zhao, B Ni, S Mei - IEEE Signal Processing Letters, 2022 - ieeexplore.ieee.org
From the perspective of probability, we propose a new method for black-box adversarial
attack via black-box variational inference (BBVI), where the knowledge of victim model is …

An Adaptive Black-box Defense against Trojan Attacks on Text Data

F Alsharadgah, A Khreishah… - … on Social Network …, 2021 - ieeexplore.ieee.org
Trojan backdoor is a poisoning attack against Neural Network (NN) classifiers in which
adversaries try to exploit the (highly desirable) model reuse property to implant Trojans into …