Identifying Audio Adversarial Examples via Anomalous Pattern Detection

Abstract

Audio processing models based on deep neural networks are susceptible to adversarial attacks even when the adversarial audio waveform is 99.9% similar to a benign sample. Given the wide application of DNN-based audio recognition systems, detecting the presence of adversarial examples is of high practical relevance. By applying anomalous pattern detection techniques in the activation space of these models, we show that 2 of the recent and current state-of-the-art adversarial attacks on audio processing systems systematically lead to higher-than-expected activation at some subset of nodes and we can detect these with up to an AUC of 0.98 with no degradation in performance on benign samples.

References in corpus (12)

• Intriguing properties of neural networks
• Higher criticism for detecting sparse heterogeneous mixtures
• Adversarial Attacks and Defences: A Survey
• Detecting Adversarial Samples from Artifacts
• Did you hear that? Adversarial Examples Against Automatic Speech Recognition
• Lingvo: a Modular and Scalable Framework for Sequence-to-Sequence Modeling
• Imperceptible, Robust, and Targeted Adversarial Examples for Automatic Speech Recognition
• Robust Audio Adversarial Example for a Physical Attack
• Houdini: Fooling Deep Structured Prediction Models
• Characterizing Audio Adversarial Examples Using Temporal Dependency
• On Fast Dropout and its Applicability to Recurrent Networks
• Isolated and Ensemble Audio Preprocessing Methods for Detecting Adversarial Examples against Automatic Speech Recognition