EUVIMEDEUVIMED entdecken
Crossref

Semantic Region-Guided Transferable Attacks on Vision-Language Pretraining Models

Jiayang Pan, Chen Wan, Wutao Chen, Lifeng Huang

IEEE Signal Processing Letters · 2026 · 33 · 2380-2384

DOI: 10.1109/lsp.2026.3701225 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Signal Processing Letters, 2026. Verfasst von Jiayang Pan und 3 weiteren Autor:innen. Die Arbeit nennt 32 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Guangdong Basic and Applied Basic Research Foundation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Vision-language pre-trained (VLP) models have achieved strong performance on multimodal tasks, but they remain vulnerable to adversarial attacks. In black-box settings, transferability is often limited because perturbation generation relies on surrogate-specific saliency cues that generalize poorly across VLP architectures. In this letter, we propose Semantic Region-Guided Attack (SRGA), a transferable attack framework that replaces such cues with detector-derived semantic regions to provide more consistent cross-model guidance. Based on these regions, SRGA performs coordinated perturbation generation in both image and text modalities with lightweight regional transformations and spatially weighted optimization. Experiments on Flickr30 K and MSCOCO show that SRGA achieves competitive transferability across diverse VLP architectures under multiple settings, with modest additional cost.

Förderung

  • National Natural Science Foundation of China
  • Guangdong Basic and Applied Basic Research Foundation

Zitieren

APA

Pan, J., Wan, C., Chen, W., & Huang, L. (2026). Semantic Region-Guided Transferable Attacks on Vision-Language Pretraining Models. IEEE Signal Processing Letters, 33, 2380-2384. https://doi.org/10.1109/lsp.2026.3701225

Vancouver

Pan J, Wan C, Chen W, Huang L. Semantic Region-Guided Transferable Attacks on Vision-Language Pretraining Models. IEEE Signal Processing Letters. 2026;33:2380-2384. doi:10.1109/lsp.2026.3701225

BibTeX
@article{Pan2026,
  author = {Pan, Jiayang and Wan, Chen and Chen, Wutao and Huang, Lifeng},
  title = {Semantic Region-Guided Transferable Attacks on Vision-Language Pretraining Models},
  journal = {IEEE Signal Processing Letters},
  year = {2026},
  volume = {33},
  pages = {2380--2384},
  doi = {10.1109/lsp.2026.3701225}
}
Kostenlos starten →