Keywords: Adversarial Machine Learning, Security
Abstract: Recent advances in adversarial machine learning have shown that defenses considered to be robust are actually susceptible to adversarial attacks which are specifically tailored to target their weaknesses. These defenses include Barrage of Random Transforms (BaRT), Friendly Adversarial Training (FAT), Trash is Treasure (TiT) and ensemble models made up of Vision Transformers (ViTs), Big Transfer models and Spiking Neural Networks (SNNs). It remains an open question, however, as to whether the adversarial examples designed to target one defense will be similarly misclassified by another defense. In this paper, we provide the first adversarial defense transferability study, as well as a game theoretic framework for ensemble adversarial attacks and defenses. Our framework is called Game theoretic Mixed Experts (GaME) and is designed to find the Mixed-Nash strategy for an attacker that can employ compositional adversarial attacks. We show that this framework creates an ensemble of defenses with greater robustness than a combinational defense with a uniform or random probability distribution. Overall, our framework and analyses advance the field of adversarial machine learning by yielding new insights into compositional attack and defense formulations.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Supplementary Material: zip
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics
Submission Guidelines: Yes
Please Choose The Closest Area That Your Submission Falls Into: Applications (eg, speech processing, computer vision, NLP)