Detecting Adversarial Samples from Artifacts
This addresses the vulnerability of DNNs to adversarial attacks, offering a detection method that is attack-agnostic, though it is incremental as it builds on existing uncertainty and feature-based techniques.
The paper tackles the problem of detecting adversarial samples in deep neural networks by using Bayesian uncertainty and density estimation in feature subspaces, achieving 85-93% ROC-AUC across various datasets and attacks.
Deep neural networks (DNNs) are powerful nonlinear architectures that are known to be robust to random perturbations of the input. However, these models are vulnerable to adversarial perturbations--small input changes crafted explicitly to fool the model. In this paper, we ask whether a DNN can distinguish adversarial samples from their normal and noisy counterparts. We investigate model confidence on adversarial samples by looking at Bayesian uncertainty estimates, available in dropout neural networks, and by performing density estimation in the subspace of deep features learned by the model. The result is a method for implicit adversarial detection that is oblivious to the attack algorithm. We evaluate this method on a variety of standard datasets including MNIST and CIFAR-10 and show that it generalizes well across different architectures and attacks. Our findings report that 85-93% ROC-AUC can be achieved on a number of standard classification tasks with a negative class that consists of both normal and noisy samples.