两个模型的故事:构建边缘模型上的规避攻击
A Tale of Two Models: Constructing Evasive Attacks on Edge Models
摘要:
全精度深度学习模型通常太大或成本太高,无法部署在边缘设备上。为了适应有限的硬件资源,使用各种边缘自适应技术(例如量化和修剪)使模型适应边缘。虽然这种技术可能对顶线准确性的影响可以忽略不计,但与原始模型相比,适应模型在输出上表现出细微的差异。在本文中,我们介绍了一种新的规避攻击,DIVA,利用这些差异的边缘适应,通过添加对抗性噪声的输入数据,最大限度地提高输出之间的差异,原始和适应模型。这种攻击特别危险,因为恶意输入将欺骗在边缘上运行的适应模型,但原始模型几乎无法检测到,原始模型通常用作权威模型版本,用于验证,调试和重新训练。我们将DIVA与最先进的攻击PGD进行了比较,结果表明DIVA在攻击适应模型时仅差1.7-3.6%,但与PGD相比,在白盒和半黑盒设置下,原始模型检测不到DIVA的可能性高1.9-4.2倍。
英文摘要:
Full-precision deep learning models are typically too large or costly to deploy on edge devices. To accommodate to the limited hardware resources, models are adapted to the edge using various edge-adaptation techniques, such as quantization and pruning. While such techniques may have a negligible impact on top-line accuracy, the adapted models exhibit subtle differences in output compared to the original model from which they are derived. In this paper, we introduce a new evasive attack, DIVA, that exploits these differences in edge adaptation, by adding adversarial noise to input data that maximizes the output difference between the original and adapted model. Such an attack is particularly dangerous, because the malicious input will trick the adapted model running on the edge, but will be virtually undetectable by the original model, which typically serves as the authoritative model version, used for validation, debugging and retraining. We compare DIVA to a state-of-the-art attack, PGD, and show that DIVA is only 1.7-3.6% worse on attacking the adapted model but 1.9-4.2 times more likely not to be detected by the the original model under a whitebox and semi-blackbox setting, compared to PGD.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。