Publication Details

Joint Energy-Based Model for Robust Speech Classification System against Dirty-Label Backdoor Poisoning Attacks

ŠŮSTEK, M.; JOSHI, S.; LI, H.; THEBAUD, T.; VILLALBA LOPEZ, J.; KHUDANPUR, S.; DEHAK, N. Joint Energy-Based Model for Robust Speech Classification System against Dirty-Label Backdoor Poisoning Attacks. Proceedings of IEEE Automatic Speech Recognition and Understanding Workshop (ASRU). Taipei: IEEE Signal Processing Society, 2023. p. 1-8. ISBN: 979-8-3503-0689-7.
Czech title
Joint Energy-Based modely pro robustní systém klasifikace řeči jako ochrana proti Dirty-Label Backdoor Poisoning útokům
Type
conference paper
Language
English
Authors
Šůstek Martin, Ing. (DCGM)
JOSHI, S.
LI, H.
THEBAUD, T.
VILLALBA LOPEZ, J.
Khudanpur Sanjeev
Dehak Najim
URL
Keywords

joint energy-based model, poisoning attacks, speech commands classification, dirty-label backdoor

Abstract

Our novel technique utilizes a Joint Energy-based Model (JEM) that integrates both discriminative and generative approaches to increase resistance against dirty-label backdoor attacks. Our approach is especially effective when the trigger is short or hardly perceivable. We simulate the attack on the Speech Commands Dataset consisting of 1 s audio clips. During training, we use JEM to model a view of the input implemented by a randomly selected 610 ms window. During inference, we combine all (40) possible views utilizing a generative part of JEM. The resulting system has slightly decreased accuracy but significantly increased resistance shown in multiple scenarios. Interestingly, replacing JEM with a standard discriminative model (Disc) provides increased resistance with a lesser effect compared to JEM but maintains accuracy. We introduce an extension motivated by semi-supervised training that further improves JEM but not Disc. JEM can also benefit from Gaussian noise during evaluation.

Published
2023
Pages
1–8
Proceedings
Proceedings of IEEE Automatic Speech Recognition and Understanding Workshop (ASRU)
ISBN
979-8-3503-0689-7
Publisher
IEEE Signal Processing Society
Place
Taipei
DOI
10.1109/ASRU57964.2023.10389697
BibTeX 
@inproceedings{BUT187975,
  author="ŠŮSTEK, M. and JOSHI, S. and LI, H. and THEBAUD, T. and VILLALBA LOPEZ, J. and KHUDANPUR, S. and DEHAK, N.",
  title="Joint Energy-Based Model for Robust Speech Classification System against Dirty-Label Backdoor Poisoning Attacks",
  booktitle="Proceedings of IEEE Automatic Speech Recognition and Understanding Workshop (ASRU)",
  year="2023",
  pages="1--8",
  publisher="IEEE Signal Processing Society",
  address="Taipei",
  doi="10.1109/ASRU57964.2023.10389697",
  isbn="979-8-3503-0689-7",
  url="https://ieeexplore.ieee.org/document/10389697"
}
Files
Back to top