Targeted Attack against Deep Neural Networks via Flipping Limited Weight Bits
This addresses a security vulnerability in deployed AI models for applications like autonomous systems, presenting a novel attack paradigm rather than an incremental improvement.
The paper tackles the problem of attacking deep neural networks by flipping a limited number of weight bits in the deployment stage to misclassify a specific sample into a target class without modifying the sample, achieving high attack success rates (e.g., over 90% on CIFAR-10) while maintaining model accuracy on other samples.
To explore the vulnerability of deep neural networks (DNNs), many attack paradigms have been well studied, such as the poisoning-based backdoor attack in the training stage and the adversarial attack in the inference stage. In this paper, we study a novel attack paradigm, which modifies model parameters in the deployment stage for malicious purposes. Specifically, our goal is to misclassify a specific sample into a target class without any sample modification, while not significantly reduce the prediction accuracy of other samples to ensure the stealthiness. To this end, we formulate this problem as a binary integer programming (BIP), since the parameters are stored as binary bits ($i.e.$, 0 and 1) in the memory. By utilizing the latest technique in integer programming, we equivalently reformulate this BIP problem as a continuous optimization problem, which can be effectively and efficiently solved using the alternating direction method of multipliers (ADMM) method. Consequently, the flipped critical bits can be easily determined through optimization, rather than using a heuristic strategy. Extensive experiments demonstrate the superiority of our method in attacking DNNs.