MetaAdvDet: Towards Robust Detection of Evolving Adversarial Attacks
This addresses the challenge of evolving adversarial attacks in machine learning security, offering a solution for scenarios where data collection is costly and attacks change frequently, though it is incremental as it builds on existing detection and meta-learning techniques.
The paper tackles the problem of detecting new adversarial attacks on deep neural networks with limited examples, proposing a meta-learning method that achieves superior performance compared to traditional detection methods on benchmarks using CIFAR-10, MNIST, and Fashion-MNIST datasets.
Deep neural networks (DNNs) are vulnerable to adversarial attack which is maliciously implemented by adding human-imperceptible perturbation to images and thus leads to incorrect prediction. Existing studies have proposed various methods to detect the new adversarial attacks. However, new attack methods keep evolving constantly and yield new adversarial examples to bypass the existing detectors. It needs to collect tens of thousands samples to train detectors, while the new attacks evolve much more frequently than the high-cost data collection. Thus, this situation leads the newly evolved attack samples to remain in small scales. To solve such few-shot problem with the evolving attack, we propose a meta-learning based robust detection method to detect new adversarial attacks with limited examples. Specifically, the learning consists of a double-network framework: a task-dedicated network and a master network which alternatively learn the detection capability for either seen attack or a new attack. To validate the effectiveness of our approach, we construct the benchmarks with few-shot-fashion protocols based on three conventional datasets, i.e. CIFAR-10, MNIST and Fashion-MNIST. Comprehensive experiments are conducted on them to verify the superiority of our approach with respect to the traditional adversarial attack detection methods.