analiza aiqbench
SpeechGuard introduces a novel run-time defense pipeline that detects and purifies poisoned audio samples in speech recognition models using adaptive perturbation injection and time-frequency masking. By autonomously generating masks based on an autoencoder, the system effectively neutralizes backdoor triggers while preserving prediction accuracy for legitimate inputs. This research addresses critical security vulnerabilities specifically within neural network architectures used for voice interaction systems.
SpeechGuard: Online Defense against Backdoor Attacks on Speech Recognition Models
arXiv:2607.15697v1 Announce Type: cross Abstract: Backdoor attacks pose a critical threat to neural network models, allowing attackers to implant a backdoor during the training phase by manipulating a small portion of the training data. In security-sensitive applications such as voice interaction for autonomous driving, the presence of backdoor attacks introduces substantial security risks. This study focuses on implementing backdoor defense measures for speech recognition models in run-time, taking into account the characteristics of audio signals. We propose SpeechGuard, the first online backdoor defense pipeline designed to identify and purify poisoned audio samples. Specifically, we improve STRIP method to perform adaptive perturbation injection to detect and filter poisoned samples, named as S-STRIP. More importantly, we further consider the purification of poisoned samples. We utilize time-frequency (T-F) masking to suppress the expression of trigger signals and autonomously generate masks based on an autoencoder. The two-stage processing prevents the backdoor in the model from being triggered, and even input speech carrying triggers can be accurately predicted. Extensive experimental demonstrate that SpeechGuard can accurately filter out poisoned samples. Through purification, it can significantly mitigate the backdoor threat while maintaining a certain prediction accuracy.
Abstract:Backdoor attacks pose a critical threat to neural network models, allowing attackers to implant a backdoor during the training phase by manipulating a small portion of the training data. In security-sensitive applications such as voice interaction for autonomous driving, the presence of backdoor attacks introduces substantial security risks. This study focuses on implementing backdoor defense measures for speech recognition models in run-time, taking into account the characteristics of audio signals. We propose SpeechGuard, the first online backdoor defense pipeline designed to identify and purify poisoned audio samples. Specifically, we improve STRIP method to perform adaptive perturbation injection to detect and filter poisoned samples, named as S-STRIP. More importantly, we further consider the purification of poisoned samples. We utilize time-frequency (T-F) masking to suppress the expression of trigger signals and autonomously generate masks based on an autoencoder. The two-stage processing prevents the backdoor in the model from being triggered, and even input speech carrying triggers can be accurately predicted. Extensive experimental demonstrate that SpeechGuard can accurately filter out poisoned samples. Through purification, it can significantly mitigate the backdoor threat while maintaining a certain prediction accuracy.
Sound (cs.SD); Cryptography and Security (cs.CR); Machine Learning (cs.LG)
arXiv:2607.15697 [cs.SD]
(or arXiv:2607.15697v1 [cs.SD] for this version)
https://doi.org/10.48550/arXiv.2607.15697 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
2024 International Joint Conference on Neural Networks (IJCNN), pp. 1-8, 2024
https://doi.org/10.1109/IJCNN60899.2024.10650300 Focus to learn more DOI(s) linking to related resources
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 305 cuvinte
Subiecte conexe
Backdoor Attacks on Speech Emotion Recognition via TTS-Generated Poisoning
arXiv · 3 sept. 2026
Backdooring Acoustic Foundation Models for Physically Realizable Triggers
arXiv · 8 oct. 2026
Beyond Waveform Robustness: Robust Feature-Vocoder Adversarial Attacks on Automatic Speech Recognition
arXiv · 6 iun. 2026
Assessing Information Ecosystems : How Governments Can Get Ahead of Hybrid Threats
carnegieendowment.org · 5 aug. 2026