Neural Architecture Dilation for Adversarial Robustness
This work addresses the trade-off between accuracy and adversarial robustness in CNNs, offering a domain-specific solution for enhancing security in machine learning applications.
The paper tackles the problem of adversarial vulnerability in convolutional neural networks (CNNs) by proposing neural architecture dilation, which improves adversarial robustness with minimal computational overhead while maintaining standard accuracy, as demonstrated on real-world datasets and benchmark networks.
With the tremendous advances in the architecture and scale of convolutional neural networks (CNNs) over the past few decades, they can easily reach or even exceed the performance of humans in certain tasks. However, a recently discovered shortcoming of CNNs is that they are vulnerable to adversarial attacks. Although the adversarial robustness of CNNs can be improved by adversarial training, there is a trade-off between standard accuracy and adversarial robustness. From the neural architecture perspective, this paper aims to improve the adversarial robustness of the backbone CNNs that have a satisfactory accuracy. Under a minimal computational overhead, the introduction of a dilation architecture is expected to be friendly with the standard performance of the backbone CNN while pursuing adversarial robustness. Theoretical analyses on the standard and adversarial error bounds naturally motivate the proposed neural architecture dilation algorithm. Experimental results on real-world datasets and benchmark neural networks demonstrate the effectiveness of the proposed algorithm to balance the accuracy and adversarial robustness.