主  讲  人： Hairong Qi 教授，美国田纳西大学

时      间： 2016年7月13日上午10:00

地      点： 信息工程学院小会议室

主 办 单 位： 必赢76net线路官网信息工程学院

主讲人介绍：Hairong Qi received the B.S. and M.S. degrees in computer science from Northern JiaoTong University, Beijing, China in 1992 and 1995， respectively, and the Ph.D. degree in computer engineering from North Carolina State University, Raleigh, in 1999. She is currently the Gonzalez Family Professor with the Department of Electrical Engineering and Computer Science at the University of Tennessee, Knoxville. Her current research interests are in advanced imaging and collaborative processing in resource-constrained distributed environment, hyperspectral image analysis, and automatic target recognition. Dr. Qi's research is supported by National Science Foundation (NSF), DARPA, Office of Naval Research (ONR), Department of Homeland Security (DHS), U.S. Army Space and Missile Defense Command, and U.S. Army Medical Research and Materiel Command. Dr. Qi is the recipient of the NSF CAREER Award. She also received the Best Paper Awards at the 18th International Conference on Pattern Recognition (ICPR) in 2006, the 3rd ACM/IEEE International Conference on Distributed Smart Cameras (ICDSC) in 2009, and IEEE Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensor (WHISPERS) in 2015. She is awarded the Highest Impact Paper from the IEEE Geoscience and Remote Sensing Society in 2012.

内 容 介 绍：This talk discusses a ubiquitous phenomenon in the physical world, i.e., everything is a mixture. Therefore, many problem solving manifests, essentially, as “unmixing”. I will start the discussion from spectral unmixing, briefly going through two unsupervised unmixing algorithms my group has developed in the past, namely, gradient-descent maximum entropy (GDME) and minimum-volume-constraint nonnegative matrix factorization (MVC-NMF). The discussion is then focused on how the concept of unmixing can be applied to solve a wide spectrum of real-world problems, including, for example, image restoration, multiple event detection, anomaly detection, and feature unmixing. The unmixing-based methodology is the key enabler to go beyond what are immediately detectable, separable, and extractable in the data, bringing unprecedented performance levels to data processing.