EUVIMEDEUVIMED entdecken
Crossref

Timestep-Compressed Attack on Spiking Neural Networks Through Timestep-Level Backpropagation

Donghwa Kang, Doohyun Kim, Sang-Ki Ko, Jinkyu Lee, Hyeongboo Baek, Brent ByungHoon Kang

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (7) · 5602-5610

DOI: 10.1609/aaai.v40i7.37479 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Donghwa Kang und 5 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

State-of-the-art (SOTA) gradient-based adversarial attacks on spiking neural networks (SNNs), which largely rely on extending FGSM and PGD frameworks, face a critical limitation: substantial attack latency from multi-timestep processing, rendering them infeasible for practical real-time applications. This inefficiency stems from their design as direct extensions of ANN paradigms, which fail to exploit key SNN properties. In this paper, we propose the timestep compressed attack (TCA), a novel framework that significantly reduces attack latency. TCA introduces two components founded on key insights into SNN behavior. First, timestep-level backpropagation (TLBP) is based on our finding that global temporal information in backpropagation to generate perturbations is not critical for an attack’s success, enabling per-timestep evaluation for early stopping. Second, adversarial membrane potential reuse (A-MPR) is motivated by the observation that initial timesteps are inefficiently spent accumulating membrane potential, a warm-up phase that can be pre-calculated and reused. Our experiments on VGG-11 and ResNet-17 with the CIFAR-10/100 and CIFAR10-DVS datasets show that TCA significantly reduces the required attack latency by up to 56.6% and 57.1% compared to SOTA methods in white-box and black-box settings, respectively, while maintaining a comparable attack success rate.

Zitieren

APA

Kang, D., Kim, D., Ko, S., Lee, J., Baek, H., & Kang, B. B. (2026). Timestep-Compressed Attack on Spiking Neural Networks Through Timestep-Level Backpropagation. Proceedings of the AAAI Conference on Artificial Intelligence, 40(7), 5602-5610. https://doi.org/10.1609/aaai.v40i7.37479

Vancouver

Kang D, Kim D, Ko S, Lee J, Baek H, Kang BB. Timestep-Compressed Attack on Spiking Neural Networks Through Timestep-Level Backpropagation. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(7):5602-5610. doi:10.1609/aaai.v40i7.37479

BibTeX
@article{Kang2026,
  author = {Kang, Donghwa and Kim, Doohyun and Ko, Sang-Ki and Lee, Jinkyu and Baek, Hyeongboo and Kang, Brent ByungHoon},
  title = {Timestep-Compressed Attack on Spiking Neural Networks Through Timestep-Level Backpropagation},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {7},
  pages = {5602--5610},
  doi = {10.1609/aaai.v40i7.37479}
}
Kostenlos starten →