Keynote Talks


Keynote Presentation Slides


Visual Computing Group, Microsoft Research Asia
Electrical and Computer Engineering, UIUC

Title: Pursuit of Low-dimensional Structures in High-dimensional Data

Abstract: In this talk, we will discuss a new class of models and techniques that can effectively model and extract rich low-dimensional structures in high-dimensional data such as images and videos, despite nonlinear transformation, gross corruption, or severely compressed measurements. This work leverages recent advancements in convex optimization for recovering low-rank or sparse signals that provide both strong theoretical guarantees and efficient and scalable algorithms for solving such high-dimensional combinatorial problems. These results and tools actually generalize to a large family of low-complexity structures whose associated regularizers are decomposable. We illustrate how these new mathematical models and tools could bring disruptive changes to solutions to many challenging tasks in computer vision and biometric recognition, such as robust recognition of faces, palm prints, and handwritten digits etc. 

This is joint work with John Wright of Columbia, Emmanuel Candes of Stanford, Zhouchen Lin of MSRA, and my students Zhengdong Zhang, Xiao Liang of Tsinghua University, Arvind Ganesh, Zihan Zhou, and Hossein Mobahi of UIUC.

Biography: Yi Ma is a Principal Researcher and the Research Manager of the Visual Computing group at Microsoft Research Asia in Beijing since January 2009. Before that he was professor at the Electrical & Computer Engineering Department of the University of Illinois at Urbana-Champaign. His main research interest is in computer vision, high-dimensional data analysis, and systems theory. He is the first author of the popular vision textbook An Invitation to 3-DVision published by Springer in 2003. Yi Ma received two Bachelors’ degree in Automation and Applied Mathematics from Tsinghua University (Beijing, China) in 1995, a Master of Science degree in EECS in 1997, a Master of Arts degree in Mathematics in 2000, and a PhD degree in EECS in 2000, all from the University of California at Berkeley. Yi Ma received the David Marr Best Paper Prize at the International Conference on Computer Vision 1999, the Longuet-Higgins Best Paper Prize at the European Conference on Computer Vision 2004, and the Sang Uk Lee Best Student Paper Award with his students at the Asian Conference on Computer Vision in 2009. He also received the CAREER Award from the National Science Foundation in 2004 and the Young Investigator Award from the Office of Naval Research in 2005. He is an associate editor of IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI) and the International Journal of Computer Vision (IJCV). He has served as the chief guest editor for special issues for the Proceedings of IEEE and the IEEE Signal Processing Magazine. He will also serve as Program Chair for ICCV 2013 and General Chair for ICCV 2015. He is a senior member of IEEE and a member of ACM, SIAM, and ASEE.



Imperial College London, Computing Department, UK
University of Twente, EEMCS, Netherlands

Title: Machine Analysis of Facial Behaviour

Abstract: Facial behaviour is our preeminent means to communicating affective and social signals. There is evidence now that patterns of facial behaviour can also be used to identify people. This talk discusses a number of components of human facial behavior, how they can be automatically sensed and analysed by computer, what is the past research in the field conducted by the iBUG group at Imperial College London, and how far we are from enabling computers to understand human facial behavior.

Biography: Maja Pantic received the M.S. and PhD degrees in computer science from Delft University of Technology, the Netherlands, in 1997 and 2001. From 2001 to 2005, she was an Assistant and then an Associate professor at Delft University of Technology, Computer Science Department. In 2006, she joined the Imperial College London, Department of Computing, UK, where she is Professor of Affective & Behavioural Computing and the leader of the iBUG group, working on machine analysis of human non-verbal behaviour. From November 2006, she also holds an appointment as the Professor of Affective & Behavioural Computing at the University of Twente, Computer Science Department, the Netherlands.

In 2007, for her research on Machine Analysis of Human Naturalistic Behavior (MAHNOB), she received European Research Council Starting Grant (ERC StG) as one of 2% best junior scientists in any research field in Europe. She is also the Scientific Director of the large European project on Social Signal Processing. In 2011, Prof. Pantic received BCS Roger Needham Award, awarded annually to a UK based researcher for a distinguished research contribution in computer science within ten years of their PhD.

She is the Editor in Chief of the Image and Vision Computing Journal (IVCJ/ IMAVIS), Associate Editor of the IEEE Transactions on Systems, Man, and Cybernetics - Part B: Cybernetics (IEEE TSMC-B), Associate Editor of the IEEE Transactions on Pattern Analysis and Machine Intelligence (IEEE TPAMI), and a member of the Steering Committee of the IEEE Transactions on Affective Computing. She is an IEEE Fellow.

Prof. Pantic is one of the world's leading experts in the research on machine understanding of human behavior including vision-based detection, tracking, and analysis of human behavioral cues like facial expressions and body gestures, and multimodal analysis of human behaviors like laughter, social signals, and affective states. She is also one of the pioneers in design and development of fully automatic, affect-sensitive human-centered anticipatory interfaces, built for humans based on human models. She has published more than 150 technical papers in the areas of machine analysis of facial expressions and emotions, machine analysis of human body gestures, and human-computer interaction. Her work is widely cited and has more than25 popular press coverage (including New Scientist, BBC Radio, and NL TV 1 and 3). See also: http://ibug.doc.ic.ac.uk/~maja/ ; http://ibug/home


Dr. Jason Pelecanos
Research Staff Member, IBM Research

Title: Automatic Speaker Recognition Research: A Journey of Recent Trends, Forensics, and Toward Achieving Robustness

Abstract: In this presentation we introduce the area of automatic speaker recognition research and discuss its strengths and remaining challenges. We share some of the interesting recent research successes from the speaker recognition community along with identifying influences from other areas of biometrics. Some of these examples include “i-vectors with PLDA” and using an appropriate measure of evidence. We then discuss what we learned by collaborating with forensic scientists and attempting to automate (largely manual) forensic voice comparison approaches. Three outcomes include event conditioned trajectory modeling, trial specific modeling, and demographics analysis. The presentation concludes with a discussion on robustness, how some prior knowledge of the signal can greatly improve performance, and what significant challenges remain.

Biography: Dr Jason Pelecanos earned his PhD at the Queensland University of Technology in 2003 in the area of signals processing. The thesis was titled “Robust Automatic Speaker Recognition”. From 2003-2006 he was a post-doctoral researcher at the IBM T.J. Watson Research Center. He is currently a Research Staff Member at IBM Research with a core focus on automatic speaker and language recognition technology. In 2009, as part of a team, he received an IBM Technical Achievement Award for advancing the field of “Conversational Biometrics”. He has also worked on and headed government research contracts in the area of automatic speaker recognition. He is actively involved in NIST speaker recognition evaluations, has served on the NSF grants review panel for speech, and is a reviewer for multiple IEEE journals and conferences. 


University of Maryland, College Park

Title: From David Marr to Samuel Johnson: Computational Vision Theories and Dictionaries for Face and Iris Recognition

Abstract: For more than five decades, numerous computational models of vision have been proposed with applications in motion and structure estimation, shape from shading, stereo, object detection and recognition. When these theories are applied to face and iris recognition/verification problems, the performance seems to degrade with variations in pose, illumination, blur and resolution, as well as changes in sensors and occlusions. During the past five years or so, it is becoming increasingly clear that sparse, discriminative, generative, linear and non-linear dictionaries built from face and iris data perform better in handling the variations mentioned above. In this talk, I will present a balanced view of the effectiveness of computational vision-based and dictionary-based approaches for face and iris recognition problems and discuss methods that integrate the best of both methodologies. The role of domain adaptation methods in realizing robust face and iris recognition systems and the link between dictionary representation and the receptive fields of neurons in V1 or striate visual cortex will also be discussed. 

Biography: Prof. Rama Chellappa received the B.E. (Hons.) degree from the University of Madras, India, in 1975 and the M.E. (Distinction) degree from Indian Institute of Science, Bangalore, in 1977. He received M.S.E.E. and Ph.D. Degrees in Electrical Engineering from Purdue University, West Lafayette, IN, in 1978 and 1981 respectively. Since 1991, he has been a Professor of Electrical Engineering and an affiliate Professor of Computer Science at University of Maryland, College Park. He is also affiliated with the Center for Automation Research (Director) and the Institute for Advanced Computer Studies (Permanent Member). In 2005, he was named a Minta Martin Professor of Engineering. Prior to joining the University of Maryland, he was an Assistant (1981-1986) and Associate Professor (1986-1991) and Director of the Signal and Image Processing Institute (1988-1990) at University of Southern California, Los Angeles. 

Over the last 29 years, he has published numerous book chapters, peer-reviewed journal and conference papers. He has co-authored and edited books on MRFs, face and gait recognition and collected works on image processing and analysis. His current research interests are face and gait analysis, markerless motion capture, 3D modeling from video, image and video-based recognition and exploitation and hyper spectral processing. Prof. Chellappa has received several awards, including an NSF Presidential Young Investigator Award, four IBM Faculty Development Awards, an Excellence in Teaching Award from the School of Engineering at USC, two paper awards from the International Association of Pattern Recognition, Society, Technical Achievement and Meritorious Service Awards from the IEEE Signal Processing Society, Technical Achievement and Meritorious Service Awards from the IEEE Computer Society. At University of Maryland, he was elected as a Distinguished Faculty Research Fellow, a Distinguished Scholar-Teacher, received the Outstanding Faculty Research Award from the College of Engineering, an Outstanding Innovator Award from the Office of Technology Commercialization and an Outstanding GEMSTONE Mentor Award. He is a Fellow of the IEEE, the International Association for Pattern Recognition and Optical Society of America.    

    Prof. Chellappa served as the associate editor of four IEEE Transactions, as a Co-Editor-in-Chief of Graphical Models and Image Processing and as an Editor-in-Chief of IEEE Transactions on Pattern Analysis and Machine Intelligence. He served as a member of the IEEE Signal Processing Society Board of Governors and as its Vice President of Awards and Membership. He is serving a two-year term as the President of IEEE Biometrics Council. He has served as a General the Technical Program Chair for several IEEE international and national conferences and workshops. He is a Golden Core Member of IEEE Computer Society and served a two-year term as a Distinguished Lecturer of the IEEE Signal Processing Society.


[Website maintained by Arun Ross and Pat Flynn]