New attack method targets adversarial transferability in vision-language models

Researchers have developed a new method called Grounding-Driven Attack (GDA) to improve the transferability of adversarial attacks against large vision-language models (LVLMs). Existing attacks often assume similar encoder architectures, but GDA focuses on text-conditioned grounding regions, which are more stable across different LVLM architectures. The proposed method allocates perturbation budgets to these grounded regions and intensifies their disruption, demonstrating superior performance in black-box scenarios. AI

IMPACT This research highlights a vulnerability in vision-language models and proposes a more effective attack strategy, potentially influencing future robustness evaluations and defense mechanisms.

RANK_REASON The cluster contains a research paper detailing a new method for adversarial attacks on large vision-language models. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• Grounding-Driven Attack
• large vision-language models
• Xinwei Zhang

• paper
• safety

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →