This work explores stylistic triggers for backdoor attacks in the audio domain: dynamic transformations of malicious samples through guitar effects. We first formalize stylistic triggers - currently missing in the literature. Second, we explore how to develop stylistic triggers in the audio domain by proposing JingleBack. Our experiments confirm the effectiveness of the attack, achieving a 96% attack success rate. Our code is available in https://github.com/skoffas/going-in-style.
@article{arxiv.2211.03117,
title = {Going In Style: Audio Backdoors Through Stylistic Transformations},
author = {Stefanos Koffas and Luca Pajola and Stjepan Picek and Mauro Conti},
journal= {arXiv preprint arXiv:2211.03117},
year = {2023}
}
Comments
Accepted to ICASSP '23 and the first two authors contributed equally