This website requires JavaScript.

Self-supervised Multi-Modal Video Forgery Attack Detection

Chenhui ZhaoXiang LiSiwen DongRabih Younes
Sep 2022
摘要
Video forgery attack threatens the surveillance system by replacing the videocaptures with unrealistic synthesis, which can be powered by the latest augmentreality and virtual reality technologies. From the machine perception aspect,visual objects often have RF signatures that are naturally synchronized withthem during recording. In contrast to video captures, the RF signatures aremore difficult to attack given their concealed and ubiquitous nature. In thiswork, we investigate multimodal video forgery attack detection methods usingboth vision and wireless modalities. Since wireless signal-based humanperception is environmentally sensitive, we propose a self-supervised trainingstrategy to enable the system to work without external annotation and thus canadapt to different environments. Our method achieves a perfect human detectionaccuracy and a high forgery attack detection accuracy of 94.38% which iscomparable with supervised methods.
展开全部
图表提取

暂无人提供速读十问回答

论文十问由沈向洋博士提出,鼓励大家带着这十个问题去阅读论文,用有用的信息构建认知模型。写出自己的十问回答,还有机会在当前页面展示哦。

Q1论文试图解决什么问题?
Q2这是否是一个新的问题?
Q3这篇文章要验证一个什么科学假设?
0
被引用
笔记
问答