RNN-Guard: Certified Robustness Against Multi-frame Attacks for Recurrent Neural Networks
This addresses a critical security vulnerability for users of RNNs in applications like speech or time-series analysis, offering a novel certified solution against multi-frame attacks.
The paper tackles the problem of defending recurrent neural networks (RNNs) against multi-frame adversarial attacks, proposing RNN-Guard as the first certified defense, which achieves up to 2.18 times higher certified robust accuracy than prior methods and improves trained models by 15.47 to 67.65 percentage points.
It is well-known that recurrent neural networks (RNNs), although widely used, are vulnerable to adversarial attacks including one-frame attacks and multi-frame attacks. Though a few certified defenses exist to provide guaranteed robustness against one-frame attacks, we prove that defending against multi-frame attacks remains a challenging problem due to their enormous perturbation space. In this paper, we propose the first certified defense against multi-frame attacks for RNNs called RNN-Guard. To address the above challenge, we adopt the perturb-all-frame strategy to construct perturbation spaces consistent with those in multi-frame attacks. However, the perturb-all-frame strategy causes a precision issue in linear relaxations. To address this issue, we introduce a novel abstract domain called InterZono and design tighter relaxations. We prove that InterZono is more precise than Zonotope yet carries the same time complexity. Experimental evaluations across various datasets and model structures show that the certified robust accuracy calculated by RNN-Guard with InterZono is up to 2.18 times higher than that with Zonotope. In addition, we extend RNN-Guard as the first certified training method against multi-frame attacks to directly enhance RNNs' robustness. The results show that the certified robust accuracy of models trained with RNN-Guard against multi-frame attacks is 15.47 to 67.65 percentage points higher than those with other training methods.