IJCB 2011 Tutorial
Sparse Representation
and Low-Rank Representation
For Biometrics
-- Theory, Algorithms,
and Applications
Description:
The recent vibrant study of sparse representation and compressive sensing has led to numerous groundbreaking results in signal processing and machine learning. In this tutorial, we will present a series of four talks to provide a high-level overview about its theory, algorithms and broad applications to biometrics. We will also point out ready-to-use MATLAB toolboxes available for participants to further acquire hands-on experience on these related topics.
13:30 Ð 14:15:
Introduction to Sparse Representation, Low-Rank Representation, and
Applications to Biometrics.
This session introduces the basic concepts of sparse representation and low-rank representation. Fundamental results about the stability and robustness will be first presented to motivate their applications in biometrics. In particular, face recognition will be used as a poster example to guide the participants through a recent high-dimensional classification framework when the biometric data may be corrupted and/or misaligned.
14:15 Ð 15:00:
L-1 minimization, entry-wise sparsity vs group sparsity, and algorithm
parallelization.
This session discusses the numerical algorithms that are responsible for recovering stable estimates of sparse signal in high-dimensional space, and provides practical solutions to biometrics applications. Two algorithms, Homotopy and Augmented Lagrangian Methods, will be first discussed to accelerate traditional L-1 minimization. Then a Lagrangian biduality framework will be introduced to derive efficient convex approximation of group sparsity constraint for structured data. Finally, we discuss how to properly implement the sparsity minimization algorithms on modern many-core CPU/GPU environments.
15:00 Ð 15:15:
Coffee Break.
15:15 Ð 16:00:
Finding structure in batches of high-dimensional data.
This session extends the techniques to enable the analysis of large batches of biometric data. We will show how tools and ideas from convex optimization give simple, robust algorithms for recovering low-rank matrices from incomplete, corrupted and noisy observations. Participants will learn how to identify problems for which these tools may be appropriate, and how to apply them effectively to solve practical problems such as robust batch image alignment and the detection of symmetric structures in images. Finally, we will show generalizations to the problem of learning sparse codes for large sets of biometric data, give example applications.
16:00 Ð 16:45: Robust
and secure iris recognition.
Iris biometric entails using the patterns on the iris as a biometric for personal authentication. It often suffers from the following three challenges: ability to handle unconstrained acquisition, privacy enhancement without compromising security and robust matching. In the fourth session, we will present a unified framework based on sparse representations and random projections that can address these issues simultaneously. Furthermore, recognition from iris videos as well as generation of cancelable iris templates for enhancing the privacy and security will be discussed.
16:45 Ð 17:00:
Conclusion and Q&A.
Speaker Bios:
Yi Ma is an associate professor (with tenure) in the Department of
Electrical
and Computer Engineering, University of Illinois at Urbana-Champaign.
He is
also an affiliated faculty of the Computer Science Department, and a
research
associate professor in the Coordinated Science Laboratory and the
Beckman
Institute. Since January 2009, he has served as a research manager for
the
Visual Computing Group at Microsoft Research Asia, Beijing. His
research
interests include computer vision and systems theory. Recent research
topics
include multiple-view geometry, vision-based control, clustering and
classification of high-dimensional data, estimation of hybrid models
and
systems. He is an associate editor of the IEEE
Transactions on
Pattern Analysis and Machine Intelligence
since 2007 and the International Journal of Computer Vision since 2010.
He has also served as the chief guest editor for special issues for the
Proceedings
of IEEE and the IEEE Signal Processing Magazine. He serves as Area
Chair for
NIPS 2011, Program Board Member for ICCV 2011, and Program Chair for
ICCV 2013.
Vishal Patel is a Research Associate at the Center
for Automation Research, University of Maryland, College
Park. He received the B.S. degrees in electrical engineering and
applied
mathematics (with honors) and the M.S. degree in applied mathematics
from North
Carolina State University, Raleigh, NC, in 2004 and 2005,
respectively. He
received the Ph.D. degree in electrical engineering from the University
of
Maryland, College Park, MD, in 2010. His research interests
include
biometrics, radar imaging, applied harmonic analysis and inverse
problems.
He is a member of Eta Kappa Nu, Pi Mu Epsilon and Phi Beta Kappa.
John Wright
is an Assistant Professor in the
Department of Electrical Engineering at Columbia University. His
research
focuses on developing provably correct and efficient tools for
recovering
low-dimensional structure in high-dimensional datasets, even when data
are
missing or grossly corrupted. These techniques address critical
estimation
problems in imaging and vision applications such as automatic face
recognition,
video stabilization and tracking, and image and data segmentation. They
also
find application outside of vision, for example in web data analysis
and
bioinformatics. He received his PhD in Electrical Engineering from the
University of Illinois at Urbana-Champaign in 2009. His work has
received a
number of awards and honors, including the 2009 Lemelson-Illinois Prize
for
Innovation for his work on robust face recognition, the 2009 UIUC
Martin Award for Excellence in Graduate Research, a 2008-2010 Microsoft
Research Fellowship, a Carver fellowship, and a UIUC Bronze Tablet
award.
Allen Y. Yang is a Research
Scientist in the Department of EECS at UC Berkeley. He has also served
as a consultant to several major companies and startups in IT industry.
His primary research areas include pattern analysis of geometric and
statistical models in very high-dimensional data spaces and
applications in motion segmentation, image segmentation, face
recognition, and signal processing in heterogeneous sensor networks. He
has published three books/chapters, ten journal papers and more than 20
conference papers. He is also the inventor of three US patents. He
received his BEng degree in Computer Science from the University of
Science and Technology of China (USTC) in 2001. From the University of
Illinois at Urbana-Champaign (UIUC), he received two MS degrees in
Electrical Engineering and Mathematics in 2003 and 2005, respectively,
and a PhD in Electrical and Computer Engineering in 2006. Among the
awards he received are a Best Bachelor's Thesis Award from USTC in
2001, a Henry Ford II Scholar Award from UIUC in 2003, a Best Paper
Award from the International Society of Information Fusion and a Best
Student Paper Award from Asian Conference on Computer Vision in 2009.