TensorShield: Tensor-based Defense Against Adversarial Attacks on Images
This work addresses security concerns in domains like image classification by providing an incremental improvement over existing defense mechanisms.
The paper tackles the problem of defending image classification systems against adversarial attacks by using tensor decomposition to discard high-frequency perturbations, resulting in a 14% improvement over the SLQ method from Shield against FGSM attacks while maintaining comparable speed.
Recent studies have demonstrated that machine learning approaches like deep neural networks (DNNs) are easily fooled by adversarial attacks. Subtle and imperceptible perturbations of the data are able to change the result of deep neural networks. Leveraging vulnerable machine learning methods raises many concerns especially in domains where security is an important factor. Therefore, it is crucial to design defense mechanisms against adversarial attacks. For the task of image classification, unnoticeable perturbations mostly occur in the high-frequency spectrum of the image. In this paper, we utilize tensor decomposition techniques as a preprocessing step to find a low-rank approximation of images which can significantly discard high-frequency perturbations. Recently a defense framework called Shield could "vaccinate" Convolutional Neural Networks (CNN) against adversarial examples by performing random-quality JPEG compressions on local patches of images on the ImageNet dataset. Our tensor-based defense mechanism outperforms the SLQ method from Shield by 14% against FastGradient Descent (FGSM) adversarial attacks, while maintaining comparable speed.