Block-wise Image Transformation with Secret Key for Adversarially Robust Defense

TL;DR

This paper introduces a block-wise image transformation technique with a secret key that significantly improves adversarial robustness in image classification, maintaining high accuracy under various attack scenarios.

Contribution

It proposes three novel algorithms for image transformation that enhance adversarial defense, outperforming existing methods on CIFAR-10 and ImageNet datasets.

Findings

Achieves near-clean image accuracy under adaptive attacks

Outperforms state-of-the-art defenses including adversarial training

Effective on both CIFAR-10 and ImageNet datasets

Abstract

In this paper, we propose a novel defensive transformation that enables us to maintain a high classification accuracy under the use of both clean images and adversarial examples for adversarially robust defense. The proposed transformation is a block-wise preprocessing technique with a secret key to input images. We developed three algorithms to realize the proposed transformation: Pixel Shuffling, Bit Flipping, and FFX Encryption. Experiments were carried out on the CIFAR-10 and ImageNet datasets by using both black-box and white-box attacks with various metrics including adaptive ones. The results show that the proposed defense achieves high accuracy close to that of using clean images even under adaptive attacks for the first time. In the best-case scenario, a model trained by using images transformed by FFX Encryption (block size of 4) yielded an accuracy of 92.30% on clean images and 91.48% under PGD attack with a noise distance of 8/255, which is close to the non-robust accuracy (95.45%) for the CIFAR-10 dataset, and it yielded an accuracy of 72.18% on clean images and 71.43% under the same attack, which is also close to the standard accuracy (73.70%) for the ImageNet dataset. Overall, all three proposed algorithms are demonstrated to outperform state-of-the-art defenses including adversarial training whether or not a model is under attack.