Presenter: Jean Gallier (Homepage)

Event Dates:
  Friday October 8, 2010 from 11:00am to 12:00pm

Computer vision problems such as contour grouping often lead to quadratic optimization problems. We begin by reviewing the problem of finding the maximum of a quadratic function, f(x) = x^T A x, on the unit sphere. Then, we describe a contour grouping problem studied by Jianbo Shi and Qihui Zhu. We present Shi's objective function (involving certain kinds of cuts in a weighted graph). Since this maximization problem is hard, we present Shi's relaxation of the problem and we show that a simpler formula can be obtained (in term of a weight matrix, P).

Presenter's Biography:

Jean Gallier is a Professor of Computer and Information Science at the University of Pennsylvania, with a secondary appointment in Mathematics. He obtained his Ph.D in Computer Science at UCLA in 1978. Gallier has done research in a number of areas of CS, including program correctness, incremental parsing and error recovery, logic programming, unification, term-rewriting and equational logic, polymorphic lambda calculi, and linear logic. For the past fifteen years, he has focused on geometric modeling, surfaces splines, surface reconstruction from meshes and recently on optimization problems arising in computer vision. Generally, he is interested in applying differential and algebraic geometry tools to problems in computer science. He is the author of four books, the last one, "Discrete Mathematics", to appear in the UTM Springer series (2011).

Presenter: Irfan Essa (Homepage)

Event Dates:
  Friday September 24, 2010 from 11:00am to 12:00pm

My research group is focused on a variety if approaches for video analysis and synthesis. In this talk, I will focus on two of our recent efforts.  One effort aimed at robust spatio-temporal segmentation of video and another on using motion and flow to predict actions from video. 

Presenter's Biography:

Irfan Essa is a Professor in the School of Interactive Computing(iC) of the College of Computing (CoC), and Adjunct Professor in the School of Electrical and Computer Engineering, Georgia Institute of Technology (GA Tech), in Atlanta, Georgia, USA.

Irfan Essa works in the areas of Computer Vision, Computer Graphics, Computational Perception, Robotics and Computer Animation, with potential impact on Video Analysis and Production (e.g., Computational Photography & Video, Image-based Modeling and Rendering, etc.) Human Computer Interaction, and Artificial Intelligence research. Specifically, he is interested in the analysis, interpretation, authoring, and synthesis (of video), with the goals of building aware environments, recognizing, modeling human activities, and behaviors, and developing dynamic and generative representations of time-varying streams. He has published over a 150 scholarly articles in leading journals and conference venues on these topics and has awards for his research and teaching.

He joined Georgia Tech Faculty in 1996 after his earning his MS (1990), Ph.D. (1994), and holding research faculty position at the Massachusetts Institute of Technology (Media Lab) [1988-1996]. His Doctoral Research was in the area of Facial Recognition, Analysis, and Synthesis.

Wednesday September 15, 2010

The GRASP Lab Quadrotor Video in Gizmodo..."The Quadrotor Drone Learns Several Terrifying New Tricks"

The last we checked in with U Penn's frighteningly maneuverable little quadrotor, we noted that it could probably zip through your window and kill you in your sleep. Well, now it can fly through windows while they're moving.....Read More

Presenter: Kingsley Fregene

Event Dates:
  Friday October 1, 2010 from 11:00am to 12:00pm

* Alternate Location: Levine 307 (3330 Walnut Street)*

This talk will describe the dynamics and control of autonomous multi-vehicle systems that are developed using the hybrid intelligent control agent (HICA) paradigm. HICA is a systems- and control-oriented approach for the modeling, control and coordination of multi-agent systems in which each agent exhibits both continuous-valued and discrete-event dynamic characteristics (i.e. is a hybrid system). A HICA essentially wraps an intelligent agent around a hybrid control system core.

Presenter's Biography:

Kingsley Fregene is a Lead Research Scientist at Lockheed Martin Advanced Technology Labs in Cherry Hill, NJ. His industrial R&D activities are in the general area of systems and control for autonomous vehicles and micro-scale devices. He received the Ph.D. and M.A.Sc. degrees from the University of Waterloo, Canada in 2002 and 1999 respectively and the B.Eng degree, with first class honors, from the Federal University of Technology, Owerri, Nigeria, in 1996 (all in Electrical & Computer Engineering).  Prior to joining Lockheed Martin, he was a Senior Research Scientist in the Guidance & Control group at Honeywell Labs, Minneapolis, where he also led several industry-university research collaborations. He has held visiting research positions at the Los Alamos and Oak Ridge National Laboratories.  He is a senior member of the IEEE and AIAA. His services to the technical community include: reviewer for the National Science Foundation Engineering Research Centers Program; program committee and associate editor for IEEE Conference on Decision and Control; IEEE Technical Committee on Aerospace Controls; and Guest editor for an upcoming special issue on UAVs and Control of the IEEE Control Systems Magazine.  He holds 2 patents, with several more pending.

Presenter: GRASP Faculty

Event Dates:
  Friday September 17, 2010 from 11:00am to 12:00pm

Sunday September 5, 2010

The 11th European Conference on Computer Vision with Program Chairs (Daniilidis, Maragos, Paragios) starts in Crete on Sep 5 with more than 700 participants. 38 oral presentatations and 287 posters were selected out of 1174 submitted.

Presenter: Robert Wood (Homepage)

Event Dates:
  Friday November 12, 2010 from 11:00am to 12:00pm

We seek to elucidate how to apply biological principles to the creation of robust, agile, inexpensive robotic insects.  However, biological inspiration alone is not sufficient to create robots that mimic the agile locomotion of their arthropod analogs.  This is particularly true as the characteristic size of the robot is decreased:  to create high performance robotic insects, we must explore novel manufacturing paradigms, develop a greater understanding of complex fluid-structure interactions for flapping-wings, generate high efficiency power and control electronics,

Presenter's Biography:

Robert Wood is an Associate Professor in Harvard's School of Engineering and Applied Sciences and a core faculty member of the Wyss Institute for Biologically Inspired Engineering. Prof. Wood completed his M.S. (2001) and Ph.D. (2004) degrees in the Dept. of Electrical Engineering and Computer Sciences at the U. C. Berkeley.  He is founder of the Harvard Microrobotics Lab which contains advanced facilities for rapid development and evaluation of unconventional robots on the micron to centimeter scale.  His current research interests involve the creation of biologically-inspired aerial and ambulatory microrobots, the unsteady aerodynamics of flapping-wing flight, minimal control of under-actuated computation-limited systems, decentralized control of multi-agent systems, artificial muscles, and morphable soft-bodied robots.  He is the winner of a 2007 DARPA Young Faculty Award, a 2008 NSF Career Award, a 2008 ONR Young Investigator Award, a 2009 Air Force Young Investigator Award, multiple best paper and best video awards, is a member of the 2008 class of Technology Review’s top 35 innovators under the age of 35, and in 2010 received the Presidential Early Career Award for Scientists and Engineers from President Obama.  Wood has served as PI or co-PI on multiple sponsored research projects including the NSF-sponsored Expeditions in Computing 'RoboBees' project which he is leading.

Presenter: Anil Jayanti Aswani

Event Dates:
  Friday August 13, 2010 from 2:00pm to 3:00pm

* Alternate Location: Levine 307 (3330 Walnut Street)*

System identification is the process of sensing the environment and using these measurements to learn the equations which describe a system. Existing approaches are unable to cope with the high-dimensionality, nonlinearity, and structure found in autonomous systems, biology, and other large networks. I will present a new statistical system identification tool which tackles some of the challenging features of these systems.

Presenter's Biography:

Anil is a postdoctoral researcher at UC Berkeley.  He received his PhD and MS in Electrical Engineering and Computer Sciences from UC Berkeley in 2010 and 2007, respectively.  Also, he earned his BS in Electrical Engineering in 2005 from the University of Michigan, in Ann Arbor.

Presenter: Kilian Pohl (Homepage)

Event Dates:
  Friday October 15, 2010 from 11:00am to 12:00pm

In this talk we develop a new curve evolution formulation for estimating the posterior distribution of objects in images. Similar to level sets, we describe the segmentation of images via a conventional likelihood model combined with a curve prior on boundaries. Unlike level sets, the curves are encoded via the logarithm-of-odds representing the posterior distribution on labels in an unconstrained vector space. The posterior distributions are sought via the Mean Field approach.

Presenter's Biography:

Kilian M Pohl received his doctorate from the Department of Computer Science at MIT and is now an Assistant Professor at the Department of Radiology, University of Pennsylvania. His main research area is computational image analysis with an emphasis on studying statistical models from a Bayesian perspective.

Monday July 26, 2010

"Australia, U.S. name finalists in worldwide robotics competition"

Six high-tech science and technology teams from four continents have been named finalists in the inaugural Multi Autonomous Ground-Robotic International Challenge, or MAGIC......They have been selected by the U.S. and the Australian departments of defense to compete this November in Australia in an effort to develop the next generation of fully-autonomous ground robots.