EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

SNEAK: Synonymous Sentences-Aware Adversarial Attack on Natural Language Video Localization

Wen Shi, Wenchao Xu, Haozhao Wang, Xingshuo Han, Pan Zhou, Ruixuan Li

ACM Transactions on Intelligent Systems and Technology · 31.10.2026 · 17 (5) · 1-26

DOI: 10.1145/3815168 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ACM Transactions on Intelligent Systems and Technology, online erschienen am 23.07.2026. Verfasst von Wen Shi und 5 weiteren Autor:innen. Die Arbeit nennt 45 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: National Key Research and Development Program of China, National Natural Science Foundation of China, Major Science and Technology Project of Hubei Province.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Natural language video localization (NLVL) is an important task in the vision-language understanding area, which calls for an in-depth understanding of not only computer vision and natural language side alone, but more importantly, the interplay between both sides. Adversarial attack has been well-recognized as a critical security issue of deep neural network models, which requires prudent investigation. Despite its extensive yet separated studies in video and language, two kinds of unimodal tasks, current understanding of the adversarial attack in NLVL tasks including both the vision and language modality is less developed. This article therefore aims to comprehensively investigate the adversarial attacks of NLVL models by examining three different facets of adversarial attack methods. To achieve the attack goal, we propose a new adversarial attack paradigm called synonymous sentences-aware adversarial attack on NLVL (SNEAK), which captures the cross-modality interplay between the vision and language sides. To further enhance the stealthiness of SNEAK, we propose a frame importance-guided pruning with SNEAK attack (PSA) mechanism to reduce the amount of perturbed frames. Extensive experiments on five NLVL models and two datasets demonstrate the effectiveness and stealthiness of the proposed attack methods.

Förderung

  • National Key Research and Development Program of China
  • National Natural Science Foundation of China
  • Major Science and Technology Project of Hubei Province
  • Hubei Science and Technology Talent Service Project
  • Ant Group through CCF-Ant Research Fund
  • HPC Platform of Huazhong University of Science and Technology

Zitieren

APA

Shi, W., Xu, W., Wang, H., Han, X., Zhou, P., & Li, R. (2026). SNEAK: Synonymous Sentences-Aware Adversarial Attack on Natural Language Video Localization. ACM Transactions on Intelligent Systems and Technology, 17(5), 1-26. https://doi.org/10.1145/3815168

Vancouver

Shi W, Xu W, Wang H, Han X, Zhou P, Li R. SNEAK: Synonymous Sentences-Aware Adversarial Attack on Natural Language Video Localization. ACM Transactions on Intelligent Systems and Technology. 2026;17(5):1-26. doi:10.1145/3815168

BibTeX
@article{Shi2026,
  author = {Shi, Wen and Xu, Wenchao and Wang, Haozhao and Han, Xingshuo and Zhou, Pan and Li, Ruixuan},
  title = {SNEAK: Synonymous Sentences-Aware Adversarial Attack on Natural Language Video Localization},
  journal = {ACM Transactions on Intelligent Systems and Technology},
  year = {2026},
  volume = {17},
  number = {5},
  pages = {1--26},
  doi = {10.1145/3815168}
}
Kostenlos starten →