all InfoSec news
Improving Adversarial Transferability of Visual-Language Pre-training Models through Collaborative Multimodal Interaction
March 19, 2024, 4:11 a.m. | Jiyuan Fu, Zhaoyu Chen, Kaixun Jiang, Haijing Guo, Jiafeng Wang, Shuyong Gao, Wenqiang Zhang
cs.CR updates on arXiv.org arxiv.org
Abstract: Despite the substantial advancements in Vision-Language Pre-training (VLP) models, their susceptibility to adversarial attacks poses a significant challenge. Existing work rarely studies the transferability of attacks on VLP models, resulting in a substantial performance gap from white-box attacks. We observe that prior work overlooks the interaction mechanisms between modalities, which plays a crucial role in understanding the intricacies of VLP models. In response, we propose a novel attack, called Collaborative Multimodal Interaction Attack (CMI-Attack), leveraging …
adversarial adversarial attacks arxiv attacks box challenge cs.cr cs.cv cs.mm gap language observe performance studies training work
More from arxiv.org / cs.CR updates on arXiv.org
Jobs in InfoSec / Cybersecurity
Sr Security Engineer - Colombia
@ Nubank | Colombia, Bogota
Security Engineer, Investigations - i3
@ Meta | Menlo Park, CA | Washington, DC | Remote, US
Cyber Security Engineer
@ ASSYSTEM | Bridgwater, United Kingdom
Security Analyst
@ Northwestern Memorial Healthcare | Chicago, IL, United States
GRC Analyst
@ Richemont | Shelton, CT, US
Security Specialist
@ Peraton | Government Site, MD, United States