报告题目：Global Perception GANs
时 间：2017年5月26日上午10:00~12:00
地 点：计算所446会议室
报告摘要:
Information theory is the mathematical theory of acquisition, representation, processing, and learning of information. In pattern recognition, information theory is the study of how pattern recognition systems accomplish these tasks. It also exploits the underlying pattern classification mechanics and new features of classical information theory. This is an introductory talk on information theory for robust pattern recognition. It will show our recent works to improve the robustness of Generative Adversarial Networks(GANs). Particularly, global and local constraints are considered in GANs to produce photo realistic/visually compelling reconstruction results for an ill-posed face reconstruction problem.
报告人简介:
Ran He received the Ph.D. degree in Pattern Recognition and Intelligent Systems from Institute of Automation, Chinese Academy of Sciences(CASIA) in 2009. He is currently a professor with National Laboratory of Pattern Recognition(NLPR), CASIA. He is also a Professor with University of Chinese Academy of Sciences. He haso btained several honors and awards such as GUCAS LuJiaxi Young Talent Award and ACM Beijing rising star award. He is currently a senior member of IEEE, serves as an associate editor of Elsevier Neurocomputing and IET imageprocessing, and serves on the program committee of several conferences. His research interests include pattern recognition, computer vision and information theory. He has published over 100 journal and conference papers in these fields, and has widely published athighly ranked international journals, such as IEEE TPAMI, IEEE TIP, IEEE TNNLS, IEEE TKDE, IEEE TIFS, Elsevier PR, and MIT NECO, and leading international conferences, such as CVPR, ICCV, AAAI, IJCAI and SIGIR.
附件下载：