Detecting Audio Attacks on ASR Systems with Dropout Uncertainty

Authors: Tejas Jayashankar, Jonathan Le Roux, Pierre Moulin
Published: 2020-06-02 | Updated: 2020-09-15

Source: https://arxiv.org/abs/2006.01906

PDF: https://arxiv.org/pdf/2006.01906

Labels Predicted by AI

敵対的攻撃検出 音声アシスタントの誤作動 攻撃タイプ

Please note that these labels were automatically added by AI. Therefore, they may not be entirely accurate.
For more details, please see the About the Literature Database page.

Abstract

Various adversarial audio attacks have recently been developed to fool automatic speech recognition (ASR) systems. We here propose a defense against such attacks based on the uncertainty introduced by dropout in neural networks. We show that our defense is able to detect attacks created through optimized perturbations and frequency masking on a state-of-the-art end-to-end ASR system. Furthermore, the defense can be made robust against attacks that are immune to noise reduction. We test our defense on Mozilla’s CommonVoice dataset, the UrbanSound dataset, and an excerpt of the LibriSpeech dataset, showing that it achieves high detection accuracy in a wide range of scenarios.