Denoised Smoothing: A Provable Defense for Pretrained Classifiers
This provides a practical solution for public vision API providers and users to enhance security against adversarial threats, though it is incremental as it builds on existing randomized smoothing techniques.
The paper tackles the problem of defending pretrained image classifiers against adversarial attacks by introducing denoised smoothing, which combines a custom-trained denoiser with randomized smoothing to guarantee robustness without modifying the classifier, achieving provable defense on datasets like ImageNet and CIFAR-10 and applying it to public APIs such as Azure and Google.
We present a method for provably defending any pretrained image classifier against $\ell_p$ adversarial attacks. This method, for instance, allows public vision API providers and users to seamlessly convert pretrained non-robust classification services into provably robust ones. By prepending a custom-trained denoiser to any off-the-shelf image classifier and using randomized smoothing, we effectively create a new classifier that is guaranteed to be $\ell_p$-robust to adversarial examples, without modifying the pretrained classifier. Our approach applies to both the white-box and the black-box settings of the pretrained classifier. We refer to this defense as denoised smoothing, and we demonstrate its effectiveness through extensive experimentation on ImageNet and CIFAR-10. Finally, we use our approach to provably defend the Azure, Google, AWS, and ClarifAI image classification APIs. Our code replicating all the experiments in the paper can be found at: https://github.com/microsoft/denoised-smoothing.