Room Impulse Responses help attackers to evade Deep Fake Detection

TL;DR

This paper demonstrates that using room impulse responses (RIRs) can significantly help fake speech evade detection systems, and proposes data augmentation with synthetic RIRs to improve robustness.

Contribution

It introduces the use of RIRs to enhance fake speech evasion and proposes data augmentation with synthetic RIRs to improve detection robustness.

Findings

RIRs significantly increase evasion success, doubling the SOTA EER.

Augmenting training data with synthetic RIRs reduces EER on reverberated fake speech.

The approach improves detection performance on both reverberated and original samples.

Abstract

The ASVspoof 2021 benchmark, a widely-used evaluation framework for anti-spoofing, consists of two subsets: Logical Access (LA) and Deepfake (DF), featuring samples with varied coding characteristics and compression artifacts. Notably, the current state-of-the-art (SOTA) system boasts impressive performance, achieving an Equal Error Rate (EER) of 0.87% on the LA subset and 2.58% on the DF. However, benchmark accuracy is no guarantee of robustness in real-world scenarios. This paper investigates the effectiveness of utilizing room impulse responses (RIRs) to enhance fake speech and increase their likelihood of evading fake speech detection systems. Our findings reveal that this simple approach significantly improves the evasion rate, doubling the SOTA system's EER. To counter this type of attack, We augmented training data with a large-scale synthetic/simulated RIR dataset. The results demonstrate significant improvement on both reverberated fake speech and original samples, reducing DF task EER to 2.13%.