Select a calendar:
Filter May Events by Event Type:
Conferences, Lectures, & Seminars
Events for May
-
CS Distinguished Lecture: Szymon Rusinkiewicz (Princeton) - Investigating the Past with 3D Scanning, Visualization, and Analysis
Thu, May 01, 2014 @ 04:00 PM - 05:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Szymon Rusinkiewicz, Princeton University
Talk Title: Investigating the Past with 3D Scanning, Visualization, and Analysis
Series: CS Distinguished Lectures
Abstract: This talk will be available to stream via the link here. [Right-click and open in new tab or window for best performance.]
Recent research into scanning, visualizing, and analyzing real-world 3D objects has the potential of providing novel insights into archaeological sites and artifacts. Two recent projects have investigated how digital methods may be used to document and propose reconstructions of objects from ancient Greece and Cyprus. The first is a system that uses 3-D and 2-D digitization hardware, together with computer-based matching techniques, to assist archaeologists and conservators in documenting and reassembling thousands of plaster fragments from wall-paintings at the site of Akrotiri (modern-day Santorini, Greece). The second is a joint research and educational project in which students created digital reconstructions of four buildings in Polis Chrysochous, Cyprus, producing a computer-animated movie to accompany an exhibition of material from the site.
Biography: Szymon Rusinkiewicz is Professor of Computer Science at Princeton University. His work focuses on the interface between computers and the visual and tangible world: acquisition, representation, analysis, and fabrication of 3D shape, motion, surface appearance, and scattering. He investigates algorithms for processing geometry and reflectance, including registration, matching, completion, hierarchical decomposition, symmetry analysis, sampling, and depiction. Applications of this work include documentation of cultural heritage artifacts and sites, appearance and performance capture for digital humans, and illustrative depiction through line drawings and non-photorealistic shading models.
Host: Hao Li
Location: SAL 101
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
CS Colloquium: Paul Elmore (Stennis Space Center) - Possiblistic & Probabilistic Approaches for Uncertainty Combinations and Measures
Wed, May 21, 2014 @ 03:00 PM - 04:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Paul Elmore, Stennis Space Center
Talk Title: Possiblistic & Probabilistic Approaches for Uncertainty Combinations and Measures
Series: CS Colloquium
Abstract: We present research on use of probabilistic and possiblistic sources of information for use in decision-making. We seek approaches to assist this process by providing information theory based quantitative evaluations to guide decisions. In particular, we examine aspects of the possiblistic conditioning of probability and analyze cases of completely certain and uncertain probability with four possibility distributions. To consider whether the conditioned probability is more informative for decision-making, three measures, Shannon entropy, Gini index and Renyi entropy are used to compare the original probability distributions and the conditioned distribution for the cases described. We found compatible results for comparing the informativeness of the original versus the conditioned probability. The three measures have increasing values with increasing uncertainty, meaning the conditioned probability will be more informative for decision making if its measure value is less than that measure for the original probability and vice-versa. Additionally, we use the Zadeh consistency measure to assess these cases and found it correlates with the evaluation results. We are investigating ways for application to actual decision-making to assess if outcomes that are more desirable result when the evaluation measures have indicated that the conditioned probability is more informative.
Biography: Dr. Paul Elmore research interests are in geospatial models and uncertainty estimation methods for geophysical information. He earned a Ph.D. in physics from University of Mississippi in 1996, where his research focus was in nonlinear acoustics. He has been a Research Physicist in the Geospatial Sciences and Technology Branch of the Naval Research Laboratory at the Stennis Space Center since 2001 and principal investigator of research and development in bathymetric data fusion and uncertainty estimation since 2007.
Host: Teamcore
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
Teamcore Seminar: Dr. Guillaume Sagnol (Zuse-Institute Berlin) - On the price of spite in Spot-checking games
Fri, May 23, 2014 @ 03:00 PM - 04:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Dr. Guillaume Sagnol, Zuse-Institute Berlin
Talk Title: On the price of spite in Spot-checking games
Series: CS Colloquium
Abstract: We introduce the class of spot-checking games (SC games). These games model problems where the goal is to distribute fare inspectors over a toll network. In an SC game, the pure strategies of network users correspond to paths in a graph, and the pure strategies of the inspectors are subset of edges to be controlled.
Although SC games are not zero-sum, we show that a (mixed) Nash equilibrium can be computed by linear programming. The computation of a strong Stackelberg equilibrium is more relevant for this problem, but we show that this is NP-hard. However, we give some bounds on the price of spite, which measures how the payoff of the inspector degrades when committing to a Nash equilibrium. That is, the inspector chooses the most harmful strategy for his opponents (because it is easy to compute) rather than the best strategy for himself. Finally, we demonstrate the quality of these bounds for a real-world application, namely the enforcement of a truck toll on German motorways.
Biography: Dr. Guillaume Sagnol is a post-doc fellow at Zuse-Institute Berlin and has been a member of this institute since 2010. He holds an engineering degree and a PhD from the French grande école "Mines Paristech". Prior to joining the ZIB he was working as a PhD student in a joint team between INRIA Saclay and Ecole Polytechnique.
His research interests lie in the areas of approximation algorithms for hard combinatorial problems, game theory, and conic programming. In particular, he has developed new SDP and SOCP formulations for the computation of "optimal experimental designs", an area at the interface of statistics and optimization. He he now working on several applied optimization projects, like the optimization of toll enforcement on German motorways, or scheduling problems in surgery planning.
Host: Teamcore
Location: Charles Lee Powell Hall (PHE) - 223
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor.