Can You Hear It?
Stefanos Koffas, Jing Xu, Mauro Conti, Stjepan Picek
Abstract
This work explores backdoor attacks for automatic speech recognition systems where we inject inaudible triggers. By doing so, we make the backdoor attack challenging to detect for legitimate users and, consequently, potentially more dangerous. We conduct experiments on two versions of a speech dataset and three neural networks and explore the performance of our attack concerning the duration, position, and type of the trigger.
Topics & Concepts
BackdoorComputer scienceArtificial neural networkComputer securityPosition (finance)Speech recognitionArtificial intelligenceDuration (music)FinanceLiteratureArtEconomicsAdversarial Robustness in Machine LearningAnomaly Detection Techniques and ApplicationsDigital Media Forensic Detection