【学术报告】：基于CNN隐写分析及对抗嵌入隐写的博弈理论框架

【报告人】：李斌

【地 点】：科技馆3楼338会议室

【时 间】：2019年4月26日（星期五） 10：00~11:00

【报告人介绍】

Dr. Bin Li received the Ph.D. degree in communication and information system from Sun Yat-sen University, Guangzhou, China, in 2009. He was a Visiting Scholar with New Jersey Institute of Technology, Newark, NJ, USA, from 2007 to 2008. He is currently a research scholar, associate professor, and Ph.D. supervisor with Shenzhen University, Shenzhen, China. He has been awarded three grants from NSFC and co-authored more than 40 IEEE journal and conference papers. Now he is the director of Shenzhen Key Laboratory of Media Security, a member of IEEE Information Forensic and Security Technical Committee (IFS-TC), vice president of Guangdong Society of Image and Graphics. His current research interests include image processing, multimedia forensics, and pattern recognition.

【报告内容简介】

  图像隐写与隐写分析是信息隐藏的主要研究方向。近年来，基于CNN的隐写分析在检测内容自适应图像隐写方面取得了不错的进展。而利用对抗样本技术，隐写方也可以采用对抗嵌入策略有效对抗基于CNN的隐写分析。通常而言，隐写分析方不能获知隐写方所采用的对抗嵌入的参数。同理，隐写方也未必能确知隐写分析方所用的参数。对于该信息的缺失，我们提出一种基于博弈理论模型的解决方案，分析隐写系统的安全性。我们的分析表明，在此模型下，非零和的博弈可以转化为零和博弈，有利于降低寻找均衡点的复杂性。此外，利用该博弈模型的结论，可以增加隐写分析检测的可靠性。

欢迎广**生参加！