Towards Query-Efficient Black-Box Adversary with Zeroth-Order Natural Gradient Descent
This addresses the need for stealthy and low-cost adversarial attacks in security-critical applications, representing an incremental improvement over existing methods.
The paper tackles the problem of query-efficient black-box adversarial attacks on deep neural networks by proposing a zeroth-order natural gradient descent method, resulting in significantly lower model query complexities compared to state-of-the-art methods, as demonstrated in empirical evaluations on image classification datasets.
Despite the great achievements of the modern deep neural networks (DNNs), the vulnerability/robustness of state-of-the-art DNNs raises security concerns in many application domains requiring high reliability. Various adversarial attacks are proposed to sabotage the learning performance of DNN models. Among those, the black-box adversarial attack methods have received special attentions owing to their practicality and simplicity. Black-box attacks usually prefer less queries in order to maintain stealthy and low costs. However, most of the current black-box attack methods adopt the first-order gradient descent method, which may come with certain deficiencies such as relatively slow convergence and high sensitivity to hyper-parameter settings. In this paper, we propose a zeroth-order natural gradient descent (ZO-NGD) method to design the adversarial attacks, which incorporates the zeroth-order gradient estimation technique catering to the black-box attack scenario and the second-order natural gradient descent to achieve higher query efficiency. The empirical evaluations on image classification datasets demonstrate that ZO-NGD can obtain significantly lower model query complexities compared with state-of-the-art attack methods.