Countering Adversarial Images Using Input Transformations
Abstract
This paper investigates strategies that defend against adversarial-example attacks on image-classification systems by transforming the inputs before feeding them to the system. Specifically, we study applying image transformations such as bit-depth reduction, JPEG compression, total variance minimization, and image quilting before feeding the image to a convolutional network classifier. Our experiments on ImageNet show that total variance minimization and image quilting are very effective defenses in practice, in particular, when the network is trained on transformed images. The strength of those defenses lies in their non-differentiable nature and their inherent randomness, which makes it difficult for an adversary to circumvent the defenses. Our best defense eliminates 60% of strong gray-box and 90% of strong black-box attacks by a variety of major attack methods.
Cite
Text
Guo et al. "Countering Adversarial Images Using Input Transformations." International Conference on Learning Representations, 2018.Markdown
[Guo et al. "Countering Adversarial Images Using Input Transformations." International Conference on Learning Representations, 2018.](https://mlanthology.org/iclr/2018/guo2018iclr-countering/)BibTeX
@inproceedings{guo2018iclr-countering,
title = {{Countering Adversarial Images Using Input Transformations}},
author = {Guo, Chuan and Rana, Mayank and Cisse, Moustapha and van der Maaten, Laurens},
booktitle = {International Conference on Learning Representations},
year = {2018},
url = {https://mlanthology.org/iclr/2018/guo2018iclr-countering/}
}