FlowHijack: A Dynamics-Aware Backdoor Attack on Flow-Matching Vision-Language-Action Models
FlowHijack: 面向流匹配视觉-语言-动作模型的动态感知后门攻击
Xinyuan An, Tao Luo, Gengyun Peng, Yaobing Wang, Kui Ren, Dongxia Wang
机构
*
Zhejiang University(浙江大学)
;
Beijing Key Laboratory of Intelligent Space Robotic Systems Technology and Applications(北京市智能空间机器人系统技术与应用重点实验室)
;
Huzhou Institute of Industrial Control Technology(湖州工业控制技术研究院)
ActDistill: General Action-Guided Self-Derived Distillation for Efficient Vision-Language-Action Models
ActDistill: 通用动作引导自衍生蒸馏用于高效视觉-语言-动作模型
Wencheng Ye, Tianshi Wang, Lei Zhu, Fengling Li, Guoli Yang, Hengtao Shen
机构
*
School of Computer Science and Technology, Tongji University(同济大学计算机科学与技术学院)
;
University of Technology Sydney(悉尼科技大学)
;
Advanced Institute of Big Data(大数据先进研究院)
机构
*
Yokohama National University(横滨国立大学)
;
Shenzhen University of Advanced Technology(深圳理工大学)
;
Zhejiang University(浙江大学)
;
University of Chinese Academy of Sciences(中国科学院大学)
;
Chengdu Institute of Computer Applications, Chinese Academy of Sciences(中国科学院成都计算机应用研究所)
;
Southeast University(东南大学)
;
University of Tsukuba(筑波大学)
;
Pusan National University(釜山大学)
;
Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences(中国科学院深圳先进技术研究院)