Towards Robust Detection of Adversarial Examples

Abstract

Although the recent progress is substantial, deep learning methods can be vulnerable to the maliciously generated adversarial examples. In this paper, we present a novel training procedure and a thresholding test strategy, towards robust detection of adversarial examples. In training, we propose to minimize the reverse cross-entropy (RCE), which encourages a deep network to learn latent representations that better distinguish adversarial examples from normal ones. In testing, we propose to use a thresholding strategy as the detector to filter out adversarial examples for reliable predictions. Our method is simple to implement using standard algorithms, with little extra training cost compared to the common cross-entropy minimization. We apply our method to defend various attacking methods on the widely used MNIST and CIFAR-10 datasets, and achieve significant improvements on robust predictions under all the threat models in the adversarial setting.

References in corpus (11)

• Delving into Transferable Adversarial Examples and Black-box Attacks
• Towards Deep Neural Network Architectures Robust to Adversarial Examples
• Characterizing Adversarial Subspaces Using Local Intrinsic Dimensionality
• Adversarial Machine Learning at Scale
• On the (Statistical) Detection of Adversarial Examples
• A Dual Approach to Scalable Verification of Deep Networks
• Dropout Inference in Bayesian Neural Networks with Alpha-divergences
• Practical Black-Box Attacks against Machine Learning
• Training verified learners with learned verifiers
• The Limitations of Deep Learning in Adversarial Settings
• Adversarial Examples Detection in Deep Networks with Convolutional Filter Statistics

Cited by in corpus (9)

• Adversarial Examples: Attacks and Defenses for Deep Learning
• Boosting Adversarial Attacks with Momentum
• Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks
• ZhuSuan: A Library for Bayesian Deep Learning
• Max-Mahalanobis Linear Discriminant Analysis Networks
• Detecting Adversarial Examples in Convolutional Neural Networks
• Improving the Robustness of Transformer-based Large Language Models with Dynamic Attention
• Detecting Adversarial Patches with Class Conditional Reconstruction Networks
• Using Anomaly Feature Vectors for Detecting, Classifying and Warning of Outlier Adversarial Examples