- talks - lists.cs.princeton.edu

C DeCoro FPO
by Melissa Lawson 20 Apr '09

20 Apr '09

Chris DeCoro will present his FPO on Monday April 27 at 2PM in Room 402. The members of his committee are: Szymon Rusinkiewicz, advisor; Tom Funkhouser and Tim Weyrich (Univ College London), readers; Fei-Fei Li and Brian Kernighan, nonreaders. A copy of his thesis, "Rendering Filters for Controlling Detail and Creating Effects", is on display in Room 310. Everyone is invited to attend his talk.

1 0

P Calamia FPO
by Melissa Lawson 20 Apr '09

20 Apr '09

Paul Calamia will present his FPO on Monday April 27 at 10AM in Room 402. The members of his committee are: Tom Funkhouser, advisor; Perry Cook and Peter Svensson (NUST), readers; Adam Finkelstein and Szymon Rusinkiewicz, nonreaders. A copy of his thesis, "Advances in Edge-Diffraction Modeling for Virtual-Acoustic Simulations" is on display in Room 310. Everyone is invited to attend his talk.

1 0

FW: Michael Littman, Rutgers, TODAY, 4:25 PM
by Ginny Hogan 20 Apr '09

20 Apr '09

The Princeton-Yahoo! machine learning speaker series continues TODAY with Michael Littman from Rutgers University. Michael is the director of the Rutgers Laboratory for Real-Life Reinforcement Learning (RL^3) and his research in machine learning examines algorithms for decision making under uncertainty. Michael will be giving a talk TODAY, at 4:25pm, in CS 105 (small auditorium). The title and abstract for his talk are included below and can also be found on the website for the Princeton-Yahoo! speaker series: http://ai.cs.princeton.edu/pmwiki/pmwiki.php?n=Public.Talks. --- Efficiently Learning to Behave Efficiently Michael Littman Rutgers University The field of reinforcement learning is concerned with the problem of learning efficient behavior from experience. In real life applications, gathering this experience is time-consuming and possibly costly, so it is critical to derive algorithms that can learn effective behavior with bounds on the experience necessary to do so. This talk presents our successful efforts to create such algorithms within a novel learning-theory framework we call "KWIK" for "knows what it knows". I'll summarize the framework, our algorithms, their formal validations, and their empirical evaluations in robotic and videogame testbeds.

1 0

M Sun general exam
by Melissa Lawson 17 Apr '09

17 Apr '09

Min Sun will present his research seminar/general exam on Wednesday April 22 at 3:30PM in Room 402. The members of his committee are; Fei-Fei Li (advisor), David Blei, Adam Finkelstein. Everyone is invited to attend his talk and those faculty wishing to remain for the oral exam following are welcome to do so. His abstract and reading list follow below. ------------------------------------------------- We propose a novel probabilistic framework for learning visual models of 3D object categories by combining appearance information and geometric constraints. Objects are represented as a coherent ensemble of parts that are consistent under 3D viewpoint transformations. Each part is a collection of salient image features. A generative framework is used for learning a model that captures the relative position of parts within each of the discretized viewpoints. Contrary to most of the existing mixture of viewpoints models, our model establishes explicit correspondences of parts across different viewpoints of the object class. Given a new image, detection and classiﬁcation are achieved by determining the position and viewpoint of the model that maximize recognition scores of the candidate objects. Our approach is among the ﬁrst to propose a generative probabilistic framework for 3D object categorization. We test our algorithm on the detection task and the viewpoint classiﬁcation task by using “car” category from both the Savarese et al. 2007 and PASCAL VOC 2006 datasets. We show promising results in both the detection and viewpoint classiﬁcation tasks on these two challenging datasets. ======= Reading list: • Textbooks [1] C. M. Bishop. Pattern Recognition and Machine Learning (Information Science and Statistics). Springer, August 2006. Chapter 1, 2, 8, and 10. [2] D. A. Forsyth and J. Ponce. Computer Vision: A Modern Approach. Prentice Hall, August 2002. • Papers [1] D. M. Blei. Variational methods for the dirichlet process. In In Proceedings of the 21st International Conference on Machine Learning, 2004. [2] P. F. Felzenszwalb and D. P. Huttenlocher. Pictorial structures for object recognition. IJCV, 61:2005, 2005. [3] D. Hoeim, C. Rother, and J. Winn. 3d layoutcrf for multi-view object class recognition and segmentation. 2007. In Proc. In IEEE Conference on Computer Vision and Pattern Recognition. [4] A. Kushal, C. Schmid, , and J. Ponce. Flexible object models for category-level 3d object recognition. 2007. Flexible object models for category-level 3d object recognition. In Proc. In IEEE Conf. on Comp. Vis. and Patt. Recogn. [5] B. Leibe, A. Leonardis, and B. Schiele. Combined object categorization and segmentation with an implicit shape model. In In ECCV workshop on statistical learning in computer vision, pages 17–32, 2004. [6] J. Liebelt, C. Schmid, and K. Schertler. Viewpointindependent object class detection using 3d feature maps. Computer Vision and Pattern Recognition, 2008. CVPR 2008. IEEE Conference on, pages 1–8, June 2008. [7] S. Savarese and L. Fei-Fei. 3d generic object categorization, localization and pose estimation. 2007. IEEE Intern. Conf. in Computer Vision (ICCV). [8] E. B. Sudderth, A. Torralba, W. T. Freeman, and A. S. Willsky. Describing visual scenes using transformed dirichlet processes. In In NIPS, pages 1297–1304. MIT Press, 2005. [9] A. Thomas, V. Ferrari, B. Leibe, T. Tuytelaars, B. Schiele, and L. V. Goo. Towards multi-view object class detection. In Proc. In IEEE Conference on Computer Vision and Pattern Recognition, volume 2, pages 1589-1596. [10] P. Yan, D. Khan, and M. Shah. 3d model based object class detection in an arbitrary view. ICCV, 2007.

1 0

S Gaw FPO
by Melissa Lawson 01 Apr '09

01 Apr '09

Shirley Gaw will present her FPO on Wednesday April 8 at 10:30AM in 306 Sherrerd Hall (note room). The members of her committee are; Ed Felten, advisor: Paul Dourish, UC Irvine and Brian Kernighan, readers; Olga Troyanskaya and Perry Cook, nonreaders. A copy of her thesis, "Ideals and Reality: Adopting Secure Technologies and Developing Secure Habits to Prevent Message Disclosure", is on display in Room 310. Everyone is invited to attend her talk.

1 0

3pm talk today on content distribution and network evolution (room 402 in CS building)
by Jennifer Rexford 30 Mar '09

30 Mar '09

Speaker: Kobus van der Merwe, AT&T Lab--Research Title: It's the Content, Stupid. How Content is driving Network Evolution. Date/time: 3:00-4:00pm on Monday March 30 Location: room 402 in the CS building Abstract: Security and robustness, operational complexity and dealing with demand growth are arguably the most significant challenges facing large scale networks. In this talk I will show how Internet growth is driven by significant increases in content related traffic. Efficiently dealing with the distribution of this content presents a significant challenge to service providers. I will present our work on load-aware anycast which offers a cross-layer approach to content distribution (considering both content and network layers) . This work is currently being readied for trial in the AT&T network and directly builds on our earlier work on Closed-loop Route Control where route selection is informed by network intelligence. As such the work presents a successful case study on the transition from research concept to production. However, this work also serves to illustrate how difficult it is to make this transition, or indeed more generally, how difficult it is to effect any change in a provider network. In the second part of my talk I will present our recent work on a platform called ShadowNet. With ShadowNet we aim to address this fundamental challenge of network change by introducing an operational trial/test network which "sits between" a lab testbed and the production network. Specifically, each ShadowNet node is composed of a collection of carrier-grade equipment (routers, switches and servers). Each node is connected to both the Internet and other ShadowNet nodes via a (virtual) backbone. ShadowNet exploits the strong separation between logical functionality and physical realization that if offered by modern equipment. As such ShadowNet is a sharable, programmable and composable infrastructure, that allows multiple technology and service trials to be executed in parallel in a realistic operational setting. I will describe the ShadowNet control framework, our prototype implementation and deployment plans. Bio: Kobus Van der Merwe is a Principal Member of Technical Staff in the Internet and Networking Systems Research Center at AT&T Labs - Research, where he is fortunate to be doing networking systems research that are both forward looking and sometimes find application in the AT&T network. His current research interests include network management, control and operation, network security and content distribution.

2 1

3pm talk today on enterprise networks in CS 402
by Jennifer Rexford 26 Mar '09

26 Mar '09

Professor Aditya Akella is visiting from U. Wisconsin-Madison on Thursday March 26, and giving a talk at 3-4pm in room 402 -- see title and abstract below. Speaker: Aditya Akella, U. Wisconsin-Madison Title: Managing the Complexity of Modern Enterprise Networks Place/time: 3-4pm Thu Mar 26 in CS 402 Abstract: Operator interviews and anecdotal evidence suggest that an operator's ability to manage a network decreases as the network becomes more complex. Today, there is no way to quantify how complex a network's design is, nor how complexity may impact network management activities. In this talk, I will present a suite of "complexity models" that describe the routing design and configuration of a network in a succinct fashion, abstracting away details of the underlying configuration languages. Our models, and the complexity metrics arising from them, capture the difficulty of configuring control and data plane behaviors on routers. They also measure the inherent complexity of the reachability constraints that a network implements via its routing design. Our models simplify network design and management by facilitating comparison between alternative designs for a network. We tested our models on seven networks, including four universities and three enterprises. We validated the results through systematic interviews with the operators of five of the networks. We found the metrics to be predictive of the issues operators face when reconfiguring their networks. A surprising result of our study was uncovering the reasons for operators choosing the designs they did. Given the frequency with which configuration errors are responsible for major outages, we believe that creating techniques to quantify the complexity of a network's design is an important first step to reducing that complexity. In databases, software engineering and other fields, metrics and benchmarks have driven the direction of the field by defining what is desirable and what is not. In proposing these metrics, we hope to start a similar conversation for network design. I will end the talk with a brief description of our recent work on WAN optimization.

1 0

Talk Today - Computer Science
by Ginny Hogan 25 Mar '09

25 Mar '09

Michelle Girvan, University of Maryland, will present her talk "Unveiling Isolated and Layered Communities in Complex Networks , at 4:15pm today Computer Science Room 105.

1 0

3pm Mon Mar 30 talk on "how content is driving network evolution" in room 402
by Jennifer Rexford 25 Mar '09

25 Mar '09

Kobus van der Merwe is visiting Monday from AT&T -- talk details below. If you'd like to visit with Kobus in the afternoon, please let me know. He has a few slots. Speaker: Kobus van der Merwe, AT&T Lab--Research Title: It's the Content, Stupid. How Content is driving Network Evolution. Location: room 402 in the CS building Date/time: 3:00-4:00pm on Monday March 30 Abstract: Security and robustness, operational complexity and dealing with demand growth are arguably the most significant challenges facing large scale networks. In this talk I will show how Internet growth is driven by significant increases in content related traffic. Efficiently dealing with the distribution of this content presents a significant challenge to service providers. I will present our work on load-aware anycast which offers a cross-layer approach to content distribution (considering both content and network layers) . This work is currently being readied for trial in the AT&T network and directly builds on our earlier work on Closed-loop Route Control where route selection is informed by network intelligence. As such the work presents a successful case study on the transition from research concept to production. However, this work also serves to illustrate how difficult it is to make this transition, or indeed more generally, how difficult it is to effect any change in a provider network. In the second part of my talk I will present our recent work on a platform called ShadowNet. With ShadowNet we aim to address this fundamental challenge of network change by introducing an operational trial/test network which "sits between" a lab testbed and the production network. Specifically, each ShadowNet node is composed of a collection of carrier-grade equipment (routers, switches and servers). Each node is connected to both the Internet and other ShadowNet nodes via a (virtual) backbone. ShadowNet exploits the strong separation between logical functionality and physical realization that if offered by modern equipment. As such ShadowNet is a sharable, programmable and composable infrastructure, that allows multiple technology and service trials to be executed in parallel in a realistic operational setting. I will describe the ShadowNet control framework, our prototype implementation and deployment plans. Bio: Kobus Van der Merwe is a Principal Member of Technical Staff in the Internet and Networking Systems Research Center at AT&T Labs - Research, where he is fortunate to be doing networking systems research that are both forward looking and sometimes find application in the AT&T network. His current research interests include network management, control and operation, network security and content distribution.

1 0

talk today @ 1:30, Gen. William Lord
by Ed Felten 25 Mar '09

25 Mar '09

General William Lord, commander of the Air Force Cyber Command, will speak today at 1:30, in the third floor open space in Sherrerd Hall. ===== Revolutions in Warfare—The Air Force’s Way Ahead When: Wednesday, March 25, 2009 - 1:30 PM Where: Sherrerd Hall, 3rd Floor Open Space Who: Major General William Lord In today’s conflicted cyber frontier, the Department of Defense and the Air Force struggle with the ways to leverage cyber for military advantage. Traditionally relegated to an Information Technology support role, cyberspace has recently emerged as a stand-alone domain, similar to land, air, sea, and space. Cyber can be viewed as an integrating domain, bringing synergy across the domains and multiplying effects in ways never before seen. Our nation’s enemies are already taking advantage of these benefits, as our nation seeks to find new ways to recruit the best and brightest to counter their actions. Amidst this backdrop, the Air Force–in partnership with other DoD and National Agencies–is forging a way forward to capitalize on these characteristics, enhancing enterprise network security and posturing capabilities to launch us into the coming decades. Bio: Maj. Gen. William T. Lord is Commander, Air Force Cyberspace Command (Provisional), Barksdale Air Force Base, La. He is responsible for establishing cyberspace as a domain in and through which the Air Force flies and fights. In his current duty, he is responsible for establishing a new numbered Air Force under Air Force Space Command which will be responsible for the organization, training and equipping of combat forces to operate in cyberspace. General Lord is a 1977 graduate of the U.S. Air Force Academy. He holds a bachelor’s degree in biological and life sciences, and master’s degrees in business administration and national resource strategy. General Lord held various duties with tours in Europe, U.S. Central Command and the White House. He has had multiple staff assignments, including two major air commands as Director of Communications and Information Systems. General Lord has commanded at the detachment, squadron, group, wing and joint levels. Prior to his current position, General Lord was Director, Cyberspace Transformation and Strategy, Secretary of the Air Force Office of Warfighting Integration and Chief Information Officer, the Pentagon, Washington, D.C.

1 0