Events Calendar

Graph Laplacian Regularizer for Inverse Imaging Problems

Speaker: Prof. Gene Cheung, National Institute of Informatics

Talk Title: Graph Laplacian Regularizer for Inverse Imaging Problems

Abstract: An inverse imaging problem is inherently an ill posed problem, and as such, appropriate definitions of image priors are crucial to regularize the problem into a well posed one. In this talk, by viewing an image patch as a graphs-signal, I describe how a graph Laplacian regularizer can be defined as a signal prior for inverse imaging problems. First, by interpreting observed graph-signals probabilistically, we show that the graph Laplacian captures the inter-pixel correlations, and the corresponding graph Fourier transform can decorrelate a graph-signal and approximate the KLT. We design a piecewise smooth image denoising algorithm based on this observation and show performance gain over state-of-the-art BM3D. Second, we show how by interpreting a graph-signal alternatively as random samples on a continuous manifold, the graph Laplacian regularizer converges to a continuous functional under mild conditions. We then show how by selecting feature functions that define the continuous gradient operator carefully, the corresponding discrete graph Laplacian regularizer is then discriminant. Based on this observation, we design a regular image denoising algorithm and bit-depth enhancement algorithm and demonstrate their merits in their respective applications.



Biography: Gene Cheung received the B.S. degree in electrical engineering from Cornell University in 1995, and the M.S. and Ph.D. degrees in electrical engineering and computer science from the University of California, Berkeley, in 1998 and 2000, respectively.

He was a senior researcher in Hewlett-Packard Laboratories Japan, Tokyo, from 2000 till 2009. He is now an associate professor in National Institute of Informatics in Tokyo, Japan.

His research interests include 3D visual representation and immersive communication. He has published over 120 international conference and ournal publications. He has served as associate editor for IEEE Transactions on Multimedia from 2007 to 2011 and currently serves as associate editor for DSP Applications Column in IEEE Signal Processing Magazine and APSIPA journal on signal & information processing, and as areaeditor for EURASIP Signal Processing: Image Communication. He currently serves as member of the Multimedia Signal Processing Technical Committee (MMSP-TC) in IEEE Signal Processing Society (2012-2014). He has also served as area chair in IEEE International Conference on Image Processing (ICIP) 2010, 2012-2013, technical program co-chair of International Packet Video Workshop (PV) 2010, track co-chair for Multimedia Signal Processing track in IEEE International Conference on Multimedia and Expo (ICME) 2011, symposium co-chair for CSSMA Symposium in IEEE GLOBECOM 2012, and area chair for ICME 2013. He was invited as plenary speaker for IEEE International Workshop on Multimedia Signal Processing (MMSP) 2013 on the topic "3D visual communication: media representation, transport and rendering'". He is a co-author of best student paper award in IEEE Workshop on Streaming and Media Communications 2011 (in conjunction with ICME 2011), best paper finalists in ICME 2011 and ICIP 2011, best paper runner-up award in ICME 2012, and best student paper award in ICIP 2013.

Host: Prof. Antonio Ortega

Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248

Audiences: Everyone Is Invited

Contact: Talyia Veal