SoK: Pitfalls in Evaluating Black-Box Attacks
Keywords: black-box attacks, adversarial attacks, attack taxonomy
TL;DR: We propose a taxonomy to categorize and better understand black-box attacks revealing unexplored threat spaces and other interesting findings, and emphasize the need to consider resource costs like attack runtime to evaluating attacks.
Abstract: Numerous works study black-box attacks on image classifiers, where adversaries generate adversarial examples against unknown target models without having access to their internal information.
However, these works make different assumptions about the adversary's knowledge, and current literature lacks cohesive organization centered around the threat model. To systematize knowledge in this area, we propose a taxonomy over the threat space spanning the axes of feedback granularity, the access of interactive queries, and the quality and quantity of the auxiliary data available to the attacker. Our new taxonomy provides three key insights.
1) Despite extensive literature, numerous under-explored threat spaces exist, which cannot be trivially solved by adapting techniques from well-explored settings. We demonstrate this by establishing a new state-of-the-art in the less-studied setting of access to \topk\ confidence scores by adapting techniques from well-explored settings of accessing the complete confidence vector but show how it still falls short of the more restrictive setting that only obtains the prediction label, highlighting the need for more research.
2) Identifying the threat models for different attacks uncovers stronger baselines that challenge prior state-of-the-art claims. We demonstrate this by enhancing an initially weaker baseline (under interactive query access) via surrogate models, effectively overturning claims in the respective paper.
3) Our taxonomy reveals interactions between attacker knowledge that connect well to related areas, such as model inversion and extraction attacks. We discuss how advances in other areas can enable stronger black-box attacks.
Finally, we emphasize the need for a more realistic assessment of attack success by factoring in local attack runtime. This approach reveals the potential for certain attacks to achieve notably higher success rates. We also highlight the need to evaluate attacks in diverse and harder settings and underscore the need for better selection criteria when picking the best candidate adversarial examples. Code is available at https://github.com/iamgroot42/blackboxsok
Submission Number: 66
Loading