Houdini: Fooling Deep Structured Prediction Models

Abstract

Generating adversarial examples is a critical step for evaluating and improving the robustness of learning machines. So far, most existing methods only work for classification and are not designed to alter the true performance measure of the problem at hand. We introduce a novel flexible approach named Houdini for generating adversarial examples specifically tailored for the final performance measure of the task considered, be it combinatorial and non-decomposable. We successfully apply Houdini to a range of applications such as speech recognition, pose estimation and semantic segmentation. In all cases, the attacks based on Houdini achieve higher success rate than those based on the traditional surrogates used to train the models while using a less perceptible adversarial perturbation.

12 pages, 8 figures, under review

Cited by in corpus (8)

• Countering Adversarial Images using Input Transformations
• POBA-GA: Perturbation Optimized Black-Box Adversarial Attacks via Genetic Algorithm
• Detecting Adversarial Attacks on Neural Network Policies with Visual Foresight
• Fooling a Real Car with Adversarial Traffic Signs
• Real-Time Adversarial Attacks
• Adversarial Security Attacks and Perturbations on Machine Learning and Deep Learning Methods
• Better the Devil you Know: An Analysis of Evasion Attacks using Out-of-Distribution Adversarial Examples
• Moving Target Defense for Deep Visual Sensing against Adversarial Examples