"Subspace Analysis, Face Detection and Recognition"

Abstract: Many problems in intelligent information processing involve some form of dimensionality reduction. A specific pattern (e.g., the human face) can reside in a low dimensional subspace of a high dimensional data (e.g., image) space.  Subspace analysis aims to discover and compute intrinsic, low dimensional features of the pattern from the data for an effective representation. In this talk, I will (1) give an overview of classic and contemporary subspace analysis techniques, (2) describe recent advances in subspace modeling and learning for face detection, alignment and recognition, and (3) show some face recognition demos.

Biography: Stan Li is a Researcher in the Microsoft Research Asian. He joined Microsoft in May 2000 from his post as an associate professor of Nanyang Technological University Singapore. His current research interest is in pattern recognition and machine learning, image analysis, and face technologies. He has published 2 books, including "Markov Random Field Modeling in Image Analysis" (Springer-Verlag, 2nd edition in 2001), and over 100 refereed papers and book chapters in these areas. He obtained a B.Eng from Hunan University, a M.Eng from National University of Defense Technology, and a PhD from Surrey University where he also worked as a research fellow. All the degrees are in Electrical and Electronic Engineering. He is a senior member of IEEE and currently serves as editorial board of Pattern Recognition, and program committees of various international conferences. He enjoys doing sports, traveling, and listening to music, especially Baroque and rock.