Towards Resistant Audio Adversarial Examples

Abstract

Adversarial examples tremendously threaten the availability and integrity of machine learning-based systems. While the feasibility of such attacks has been observed first in the domain of image processing, recent research shows that speech recognition is also susceptible to adversarial attacks. However, reliably bridging the air gap (i.e., making the adversarial examples work when recorded via a microphone) has so far eluded researchers. We find that due to flaws in the generation process, state-of-the-art adversarial example generation methods cause overfitting because of the binning operation in the target speech recognition system (e.g., Mozilla Deepspeech). We devise an approach to mitigate this flaw and find that our method improves generation of adversarial examples with varying offsets. We confirm the significant improvement with our approach by empirical comparison of the edit distance in a realistic over-the-air setting. Our approach states a significant step towards over-the-air attacks. We publish the code and an applicable implementation of our approach.

References in corpus (16)

• Intriguing properties of neural networks
• Towards Deep Learning Models Resistant to Adversarial Attacks
• Evasion Attacks against Machine Learning at Test Time
• Synthesizing Robust Adversarial Examples
• On Evaluating Adversarial Robustness
• DolphinAtack: Inaudible Voice Commands
• Adversarial Examples: Opportunities and Challenges
• Natural Adversarial Examples
• Lingvo: a Modular and Scalable Framework for Sequence-to-Sequence Modeling
• Imperceptible, Robust, and Targeted Adversarial Examples for Automatic Speech Recognition
• Why do deep convolutional networks generalize so poorly to small image transformations?
• Robust Audio Adversarial Example for a Physical Attack
• Adversarial Examples that Fool Detectors
• Characterizing Audio Adversarial Examples Using Temporal Dependency
• Inaudible Voice Commands
• SirenAttack: Generating Adversarial Audio for End-to-End Acoustic Systems

Cited by in corpus (3)

• Visualizing Automatic Speech Recognition -- Means for a Better Understanding?
• Can audio-visual integration strengthen robustness under multimodal attacks?
• An Integrated Algorithm for Robust and Imperceptible Audio Adversarial Examples