Defending Video Recognition Model Against Adversarial Perturbations via Defense Patterns
IEEE Transactions on Dependable and Secure Computing · 2024-07 · 21 (4) · 4110-4121
DOI: 10.1109/tdsc.2023.3346064 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Dependable and Secure Computing, 2024-07. Verfasst von Hong Joo Lee und 1 weiteren Autor:innen. Die Arbeit wurde bisher 7-mal zitiert (Crossref). Gefördert von: Connected AI Education & Research Program for Industry and Society Innovation, Center for Applied Research in Artificial Intelligence.
Abstract (Originalsprache)
Deep Neural Networks (DNNs) have been widely successful in various domains, but they are vulnerable to adversarial attacks. Recent studies have also demonstrated that video recognition models are susceptible to adversarial perturbations, but the existing defense strategies in the image domain do not transfer well to the video domain due to the lack of considering temporal development and require a high computational cost for training video recognition models. This article, first, investigates the temporal vulnerability of video recognition models by quantifying the effect of temporal perturbations on the model's performance. Based on these investigations, we propose Defense Patterns (DPs) that can effectively protect video recognition models by adding them to the input video frames. The DPs are generated on top of a pre-trained model, eliminating the need for retraining or fine-tuning, which significantly reduces the computational cost. Experimental results on two benchmark datasets and various action recognition models demonstrate the effectiveness of the proposed method in enhancing the robustness of video recognition models.
Förderung
- Connected AI Education & Research Program for Industry and Society Innovation
- Center for Applied Research in Artificial Intelligence
Zitieren
APA
Lee, H. J., & Ro, Y. M. (2024). Defending Video Recognition Model Against Adversarial Perturbations via Defense Patterns. IEEE Transactions on Dependable and Secure Computing, 21(4), 4110-4121. https://doi.org/10.1109/tdsc.2023.3346064
Vancouver
Lee HJ, Ro YM. Defending Video Recognition Model Against Adversarial Perturbations via Defense Patterns. IEEE Transactions on Dependable and Secure Computing. 2024;21(4):4110-4121. doi:10.1109/tdsc.2023.3346064
BibTeX
@article{Lee2024,
author = {Lee, Hong Joo and Ro, Yong Man},
title = {Defending Video Recognition Model Against Adversarial Perturbations via Defense Patterns},
journal = {IEEE Transactions on Dependable and Secure Computing},
year = {2024},
volume = {21},
number = {4},
pages = {4110--4121},
doi = {10.1109/tdsc.2023.3346064}
}
EUVIMED