Detecting Audio Attacks on ASR Systems with Dropout Uncertainty

Understand

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.