arXiv preprint / cs.CV

Two Sides of the Same Coin: Co-Evolving Search for Cross-Task Attacks on Vision-Language Models

作者：Xuanhui Lin, Junhao Dong, Mingrong Gong, Yucheng Chen, Xinghua Qu, Yew-Soon Ong · 会议/期刊：arXiv preprint · 方向：cs.CV · 发布日期：2026-08-04

2026-08-04视觉语言视觉感知
新标签打开 AlphaXiv
Two Sides of the Same Coin: Co-Evolving Search for Cross-Task Attacks on Vision-Language Models figure
AlphaXiv 中文概览（可滚动查看）