When the Differences in Frequency Domain are Compensated: Understanding and Defeating Modulated Replay Attacks on Automatic Speech Recognition

TL;DR

This paper reveals a new modulated replay attack on ASR systems that bypasses existing defenses, and proposes DualGuard, a joint time-frequency domain detection method, achieving 98% accuracy in real-world tests.

Contribution

The paper introduces the concept of modulated replay attacks and develops DualGuard, a novel detection system that effectively counters these attacks by analyzing both time and frequency domain features.

Findings

Modulated replay attacks can bypass existing frequency domain defenses.

DualGuard detects replay attacks with 98% accuracy in real-world experiments.

Replay attacks leave detectable artifacts in either time or frequency domain.

Abstract

Automatic speech recognition (ASR) systems have been widely deployed in modern smart devices to provide convenient and diverse voice-controlled services. Since ASR systems are vulnerable to audio replay attacks that can spoof and mislead ASR systems, a number of defense systems have been proposed to identify replayed audio signals based on the speakers' unique acoustic features in the frequency domain. In this paper, we uncover a new type of replay attack called modulated replay attack, which can bypass the existing frequency domain based defense systems. The basic idea is to compensate for the frequency distortion of a given electronic speaker using an inverse filter that is customized to the speaker's transform characteristics. Our experiments on real smart devices confirm the modulated replay attacks can successfully escape the existing detection mechanisms that rely on identifying suspicious features in the frequency domain. To defeat modulated replay attacks, we design and implement a countermeasure named DualGuard. We discover and formally prove that no matter how the replay audio signals could be modulated, the replay attacks will either leave ringing artifacts in the time domain or cause spectrum distortion in the frequency domain. Therefore, by jointly checking suspicious features in both frequency and time domains, DualGuard can successfully detect various replay attacks including the modulated replay attacks. We implement a prototype of DualGuard on a popular voice interactive platform, ReSpeaker Core v2. The experimental results show DualGuard can achieve 98% accuracy on detecting modulated replay attacks.