SUNMONTUEWEDTHUFRISAT
Conferences, Lectures, & Seminars
Events for August
-
CS Colloq: Prof. Simon Lucey
Tue, Aug 18, 2009 @ 11:00 AM - 12:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Time: 11 AM - 12:30 PMLocation: PHE 333Talk title: Leveraging Convex Quadratics for Alignment and Classification in Vision
Speaker: Prof. Simon Lucey (Robotics Institute, CMU)
Host: Prof. Louis-Philippe MorencyAbstract:
Convex quadratic objective functions are an attractive means of expressing some goal/task in vision like alignment or classification as: (i) local minima = global minimum, (ii) the sum of N convex quadratics is itself a convex quadratic, and (iii) they offer computationally efficient solutions. Famous algorithms in vision such as the Lucas-Kanade (LK) algorithm (alignment), and the Support Vector Machines (SVM) can both be viewed as minimizing a convex quadratic objective function.In the first part of this talk, I will be exploring problems/applications in vision where the original objective function can be suitably relaxed to take advantage of the convex quadratic form.
We will introduce two examples of such successful relaxations: (i) Convex Quadratic Fitting (CQF) for non-rigid face alignment with local-experts, and (ii) Least-Squares Congealing (LSC) for the task of unsupervised image ensemble alignment. Both examples, at the time of writing, exhibit superior performance to current state of the art performance.In the second part of this talk, I will explore the concept that if our learning goal can be expressed as a convex quadratic, and our feature extraction step linear, then the sequential feature extraction and optimization steps can be re-interpreted within a single learning goal. This alternate view of linear feature extraction with respect to a convex quadratic learning goal has a number of advantages. First, for the case of classification within the well known linear support vector machine (SVM) framework the memory and computational overheads, typically occurring due to the high dimensionality of the feature extraction process now disappear. From a theoretical perspective the feature extraction step can now be viewed alternately as manipulating the margin of the SVM. This insight is synergetic with recent work in learning that has demonstrated that the choice of margin employed while learning a SVM is critical for high classification performance in many circumstances. Second, for the case of alignment we demonstrate that a similar approach can be applied when employing linear feature extraction in conjunction with the LK algorithm for alignment. By framing the LK algorithm within the Fourier-domain, an algorithm we refer to as Fourier-LK (FLK), we demonstrate superior alignment performance with nearly no additional computational overhead.Bio:
Simon Lucey is an Assistant Research Professor in the Robotics Institute at Carnegie Mellon University, and has been a faculty member there since October 2005. Before that he was a Post-Doc in the Electrical and Computer Engineering (ECE) department at Carnegie Mellon University. Dr. Lucey's research interests are in computer vision, pattern recognition and machine learning with specific interests in their application to space-time face and body analysis. He received his Ph.D. in 2003 on the topic of audio-visual speaker and speech recognition from the Queensland University of Technology (QUT), Australia. To his credit he has over 30 publications in international conferences, journals and book chapters. He has been a reviewer for a number of international journals and conferences in vision, learning, pattern recognition and multimedia. He has organized and co-chaired a number of conferences, workshops and special sessions, including last year's International Conference on Auditory and Visual Speech Processing(AVSP'08) and the successful "Beyond Patches" workshop series at CVPR'06 and CVPR'07. His work on face tracking and recognition was recently showcased on a Discovery Channel series "Weird Connections". Simon has served on the programme committee for a number of top international computer vision and pattern recognition conferences including CVPR, ICCV, ECCV and BMVC and also served as an Associate Editor for the IEEE Transactions of Multimedia.
Location: Charles Lee Powell Hall (PHE) - 333
Audiences: Everyone Is Invited
Contact: CS Front Desk
-
CS Colloq: Dr. Janusz Marecki
Mon, Aug 24, 2009 @ 03:00 PM - 04:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Time: 3 PM - 4:30 PMLocation: VKC 150Talk title: "A Quest for Intelligence: From Stochastic Optimization to Cortex Simulations"
Speaker: Dr. Janusz Marecki (IBM T.J. Watson Research Center)
Host: Prof. Milind TambeAbstract:
With the computational capacity of the world's supercomputers approaching the computational capacity of the human brain, we may be only at the beginning of an era of intelligent algorithms. In this talk I will summarize my quest for intelligent algorithms, focused on
(i) developing techniques for risk-sensitive, autonomous decision making in stochastic environments and (ii) deciphering the cortical algorithm and running it on massively-parallel architectures, to simulate the perception/cognition functions of the brain. I will conclude by proposing a set of measures, to make the deployed intelligent algorithms safer.Bio:
Janusz Marecki is currently a Research Staff Member in the Mathematical Sciences Department at IBM T.J. Watson Research Center and a visiting professor at the Academy of Computer Science in Poland.
He holds a PhD in Artificial Intelligence from University of Southern California and a DrSc in Mathematical Modeling from Institute of Information Infrastructure in Lviv. Janusz's research interests include decision/game theory, multiagent systems, neural networks and high-performance computing. He is an author of over 60 refereed publications and 2 patents.
(More information at http://teamcore.usc.edu/marecki)
Location: Von Kleinsmid Center For International & Public Affairs (VKC) - 150
Audiences: Everyone Is Invited
Contact: CS Front Desk
