Identification of Attack-Specific Signatures in Adversarial Examples

Abstract

The adversarial attack literature contains a myriad of algorithms for crafting perturbations which yield pathological behavior in neural networks. In many cases, multiple algorithms target the same tasks and even enforce the same constraints. In this work, we show that different attack algorithms produce adversarial examples which are distinct not only in their effectiveness but also in how they qualitatively affect their victims. We begin by demonstrating that one can determine the attack algorithm that crafted an adversarial example. Then, we leverage recent advances in parameter-space saliency maps to show, both visually and quantitatively, that adversarial attack algorithms differ in which parts of the network and image they target. Our findings suggest that prospective adversarial attacks should be compared not only via their success rates at fooling models but also via deeper downstream effects they have on victims.

References in corpus (11)

• Adam: A Method for Stochastic Optimization
• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Two-Stream Convolutional Networks for Action Recognition in Videos
• Towards Deep Learning Models Resistant to Adversarial Attacks
• Theoretically Principled Trade-off between Robustness and Accuracy
• Detecting Adversarial Samples from Artifacts
• On the (Statistical) Detection of Adversarial Examples
• LaVAN: Localized and Visible Adversarial Noise
• Investigating the significance of adversarial attacks and their relation to interpretability for radar-based human activity recognition systems
• LowKey: Leveraging Adversarial Attacks to Protect Social Media Users from Facial Recognition
• Where do Models go Wrong? Parameter-Space Saliency Maps for Explainability