Select a calendar:
Filter April Events by Event Type:
SUNMONTUEWEDTHUFRISAT
Conferences, Lectures, & Seminars
Events for April
-
CS Student Colloquium: Zhenzhen Gao - City-Scale Aerial LiDAR Point Cloud Visualization
Thu, Apr 03, 2014 @ 04:00 PM - 05:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Zhenzhen Gao, USC
Talk Title: City-Scale Aerial LiDAR Point Cloud Visualization
Series: Student Seminar Series
Abstract: Aerial LiDAR (Light Detection and Ranging) is cost-effective in acquiring terrain and urban information by mounting a downward-scanning laser on a low-flying aircraft. It produces huge volumes of unconnected 3D points. This thesis focuses on the interactive visualization of aerial LiDAR point clouds of cities, which is applicable to a number of areas including virtual tourism, security, land management and urban planning.
A framework needs to address several challenges in order to deliver useful visualizations of aerial LiDAR cities. Firstly, the data is 2.5D, in that the sensor is only able to capture dense details of the surfaces facing it, leaving few samples on vertical building walls. Secondly, the data often suffers from noise and under-sampling. Finally, the large size of the data can easily exceed the memory capacity of a computer system.
This thesis first introduces a visually-complete rendering framework for aerial LiDAR cities. By inferring classification information, building walls and occluded ground areas under tree canopies are completed either through pre-processing point cloud augmentation or through online procedural geometry generation. A multi-resolution out-of-core strategy and GPU-accelerated rendering enable interactive visualization of virtually unlimited size data. With adding only a slight overhead to existing point-based approaches, the framework provides comparable quality to visualizations of off-line pre-computation of 3D polygonal models.
The thesis then presents a scalable out-of-core algorithm for mapping colors from aerial oblique imagery to city-scale aerial LiDAR points. Without intensive processing of points, colors are mapped via a modified visibility pass of GPU splatting, and a weighting scheme leveraging image resolution and surface orientation.
To alleviate visual artifacts caused by noise and under-sampling, the thesis shows an off-line point cloud refinement algorithm. By explicitly regularizing building boundary points, the algorithm can effectively remove noise, fill gaps, and preserve and enhance both normal and position discontinuous features for piece-wise smoothing buildings with arbitrary shape and complexity.
Finally, the thesis introduces a new multi-resolution rendering framework that supports real-time refinement of aerial LiDAR cities. Without complex computation and without user interference, simply based on curvature analysis of points of uniform sized spatial partitions, hierarchical hybrid structures are constructed indicating whether to represent a partition as point or polygon. With the help of such structures, both rendering and refinement are dynamically adaptive to views and curvatures. Compared to visually-complete rendering, the new framework is able to deliver comparable visual quality with less than 8\% increase in pre-processing time and 2-5 times higher rendering frame-rates. Experiments on several cities show that the refinement improves rendering quality for large magnification under real-time constraint.
Host: CS PHD Committee
Location: Henry Salvatori Computer Science Center (SAL) - 101
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Abhishek Jain (MIT CSAIL): Computing on Private Data
Tue, Apr 08, 2014 @ 04:00 PM - 05:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Abhishek Jain, MIT CSAIL
Talk Title: Computing on Private Data
Series: CS Colloquium
Abstract: THIS TALK WILL BE BROADCAST / STREAMING VIA THE FOLLOWING LINK. (Right click-open link in new tab or window.)
Today, end users generate large volumes of private data, some of which may be stored on the cloud in an encrypted form. The need to perform computation on this data to extract meaningful information has become ubiquitous.
The following fundamental questions arise in this setting: Can the cloud compute on the encrypted data of multiple users without knowing their secret keys? What functions can be computed in this manner? What if the users are mutually distrustful?
My research provides the first positive resolution of these questions. In this talk I will describe these new results and my other interests.
Biography: Abhishek Jain is currently a postdoctoral researcher in the Cryptography and Information Security Group at MIT CSAIL and Boston University. He received his PhD in Computer Science from UCLA in 2012 where he was the recipient of the Symantec Outstanding Graduate Student Research award. Abhishek's research interests are in cryptography and security, and related areas of theoretical computer science.
Host: Ming-Deh Huang
Location: Henry Salvatori Computer Science Center (SAL) - 101
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Nitin Agrawal (NEC Labs Princeton ) - Rethinking Data Abstractions for Mobile Apps
Thu, Apr 10, 2014 @ 11:00 AM - 12:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Nitin Agrawal, NEC Labs Princeton
Talk Title: Rethinking Data Abstractions for Mobile Apps
Series: CS Colloquium
Abstract: Mobile apps have radically changed the ways in which users store, interact, and share data. A crucial component, for building high-quality mobile apps, nowadays is the infrastructure for managing data â both locally on mobile devices and remotely through cloud-based services. In building such “data-centric” mobile apps, developers benefit from several abstractions available for local and remote I/O. In this talk, I will present evidence as to why existing data abstractions, for local storage, are counter-productive for performance, and for cloud sync, are insufficient for consistency, efficiency, and programmability. As part of our work we are rethinking the data abstractions that will empower app developers to write and deploy such apps with ease. I will present a novel data-management platform, Simba, which provides a powerful yet easy-to-use API for mobile data storage and sync. Using Simba, apps take significantly less effort to write, compared to commercially-available sync services like Dropbox, while being more efficient.
Biography: Nitin Agrawal works as a Researcher in the Storage Systems group at NEC Labs Princeton after graduating with a PhD from Wisconsin in 2009. His interests lie in distributed and mobile systems, operating systems, applied machine learning, and storage systems, and his recent research focuses on cloud infrastructure for data-centric mobile services. He has received Best Paper Awards at FAST 2009, FAST 2011, FAST 2012, and a top paper selection at FAST 2007. More details can be found at http://www.nec-labs.com/~nitin/
Host: Ramesh Govindan
Location: Henry Salvatori Computer Science Center (SAL) - 322
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Aleksandra Korolova (Google) - Scalable Algorithms for Protecting User Privacy
Thu, Apr 10, 2014 @ 04:00 PM - 05:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Aleksandra Korolova, Google
Talk Title: Scalable Algorithms for Protecting User Privacy
Series: CS Colloquium
Abstract: THIS TALK WILL BE BROADCAST / STREAMING VIA THE FOLLOWING LINK. (Right click-open link in new tab or window.)
Ubiquitous use of the Internet and mobile technologies combined with dropping data storage and processing costs have enabled new forms of communications and data-driven innovations. However, they have also created unprecedented challenges for privacy, with companies, policy makers, and individuals struggling in their search for approaches that could enable innovation while avoiding privacy harms.
In this talk, I will present algorithmic and data-mining research that demonstrates how these seemingly conflicting goals may be achieved, even when the data being collected about individuals is constantly changing and expanding. I will first demonstrate that merely restricting data sharing is insufficient to protect privacy via a novel attack exploiting Facebook's ad targeting system to reveal users’ secrets. I will then present algorithms that enable useful search data releases and social advertising computations while provably protecting privacy. Finally, I will show how data mining techniques used to improve web search and advertising quality can be effectively applied towards improving privacy policies and building tools for safer user experiences.
Biography: Aleksandra Korolova is a research scientist at Google, where she works on developing and implementing approaches for privacy-preserving data mining and for data-driven understanding of user privacy preferences. Aleksandra received her Ph.D. in Computer Science from Stanford, where she was a Cisco Systems Stanford Graduate Fellow. Aleksandra's thesis, "Protecting Privacy when Mining and Sharing User Data", was awarded the Arthur L. Samuel Award for the best 2011-2012 CS Ph.D. thesis at Stanford, and her work on "Privacy Violations Using Microtargeted Ads" was a co-winner of the 2011 PET Award for Outstanding Research in Privacy Enhancing Technologies. Prior to joining Google, Aleksandra has interned at PARC, Facebook, Microsoft, and Yahoo! Research. She graduated Phi Beta Kappa from MIT with a B.S. degree in Mathematics with Computer Science.
Host: Shanghua Teng
Location: Henry Salvatori Computer Science Center (SAL) - 101
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
Second USC/VSoE Symposium on the Futures of Robotics
Tue, Apr 15, 2014 @ 09:00 AM - 05:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Multiple, Multiple
Talk Title: Second USC / VSoE Symposium on the Futures of Robotics
Series: CS Symposium Series
Abstract: The second USC Symposium on the Futures of Robotics will be held on the USC campus (Davidson Conference Center [DCC] Club Suite AB) on April 15, 2014. The academic community is cordially invited. No registration is necessary.
The symposium is a day long set of talks by young- to mid-career roboticists breaking new ground in emerging areas in robotics and related fields.
This event will be held on:
Tuesday, April 15, 2014
9:00am - 5:00pm
Agenda
08:30 - 09:00 Breakfast and Greetings
09:00 - 09:10 Gaurav Sukhatme, Maja Matarić, Stefan Schaal, Nora Ayanian - Welcoming remarks
09:15 - 09:50 Andrea Thomaz, Georgia Institute of Technology
- "Social Robot Learning"
09:55 - 10:30 Julie Shah, Massachusetts Institute of Technology
- "Integrating Robots into Team-Oriented Environments"
10:35 - 10:50 Break
10:55 - 11:30 Stephanie Gil, CSAIL, MIT
- "Adaptive Communication Networks for Heterogeneous Teams of Robots"
11:35 - 12:10 M. Ani Hsieh, Drexel University
- "Control and Coordination of Robot Teams in Geophysical Flows: Exploiting the Environment for Prolonged Autonomy"
12:15 - 13:30 Lunch - University Club [By Invitation Only]
13:35 - 14:10 Cynthia Sung, CSAIL, MIT
- "Geometric Design of Print-and-Fold Robots via Composition"
14:15 - 14:50 Sonia Chernova, Worcester Polytechnic Institute
- "Crowds and Robots: Enabling Robots to Learn from Everyday People"
14:55 - 15:10 Break
15:15 - 15:55 Brenna Argall, Northwestern University
- "Turning Assistive Machines into Assistive Robots"
15:55 - 16:30 Anca Dragan, Carnegie Mellon University
- "A Mathematical Formalism for Legible Robot Motion"
16:35 - 16:40 Walk to labs
16:40 - 18:00 USC Robotics labs tours [Ronald Tutor Hall, 4th Floor]
18:00 - 18:30 Travel to dinner
18:30 - 22:00 Dinner (by Invitation only)
Perch [448 S Hill St, Los Angeles, CA, 90013, 213-802-1770]
Event Location
The Davidson Continuing Education Conference Center
University of Southern California, 3415 South Figueroa Street
Los Angeles, CA 90089-0871
Contact Us
Jacob Beal
Event Coordinator
jbeal@usc.edu
213-740-4498
Biography: More details, lecture abstracts and biographies can be found at the dedicated page here.
Host: Gaurav Sukhatme
More Info: http://www.cs.usc.edu/research/2nd-usc-symposium-futures-of-robotics.htm
More Information: ROBOTICS_Fullsheet.jpg
Location: Charlotte S. & Davre R. Davidson Continuing Education Conference Center (DCC) - Club AB
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
Event Link: http://www.cs.usc.edu/research/2nd-usc-symposium-futures-of-robotics.htm
-
CS Student Colloquium: Benjamin Ford - Adaptive Resource Allocation for Wildlife Protection against Illegal Poachers & Thanh H. Nguyen - Stop the Compartmentalization: Unified Robust Algorithms for Handling Uncertainties in Security Games
Tue, Apr 15, 2014 @ 04:00 PM - 05:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Benjamin Ford, Thanh H. Nguyen, USC CS
Talk Title: Benjamin Ford - Adaptive Resource Allocation for Wildlife Protection against Illegal Poachers & Thanh H. Nguyen - Stop the Compartmentalization: Unified Robust Algorithms for Handling Uncertainties in Security Games
Series: Student Seminar Series
Abstract: Benjamin Ford - Adaptive Resource Allocation for Wildlife Protection against Illegal Poachers
Abstract: Illegal poaching is an international problem that leads to the extinction of species and the destruction of ecosystems. As evidenced by dangerously dwindling populations of endangered species, existing anti-poaching mechanisms are insufficient. This paper introduces the Protection Assistant for Wildlife Security (PAWS) application - a joint deployment effort done with researchers at Uganda’s Queen Elizabeth National Park (QENP) with the goal of improving wildlife ranger patrols. While previous works have deployed applications with a game-theoretic approach (specifically Stackelberg Games) for counter-terrorism, wildlife crime is an important domain that promotes a wide range of new deployments. Additionally, this domain presents new research challenges and opportunities related to learning behavioral models from collected poaching data. In addressing these challenges, our first contribution is a behavioral model extension that captures the heterogeneity of poachers’ decision making processes. Second, we provide a novel framework, PAWS-Learn, that incrementally improves the behavioral model of the poacher population with more data. Third, we develop a new algorithm, PAWS-Adapt, that adaptively improves the resource allocation strategy against the learned model of poachers. Fourth, we demonstrate PAWS’s potential effectiveness when applied to patrols in QENP, where PAWS will be deployed.
Thanh H. Nguyen - Stop the Compartmentalization: Unified Robust Algorithms for Handling Uncertainties in Security Games
Given the real-world applications of Stackelberg security games (SSGs), addressing uncertainties in these games is a major challenge. Unfortunately, we lack any unified computational framework for handling uncertainties in SSGs. Current state-of-the-art has provided only compartmentalized robust algorithms that handle uncertainty exclusively either in the defender’s strategy or in adversary’s payoff or in the adversary’s rationality, leading to potential failures in real-world scenarios where a defender often faces multiple types of uncertainties. Furthermore, insights for improving performance are not leveraged across the compartments, leading to significant losses in quality or efficiency. In this paper, we provide the following main contributions: 1) we present the first unified framework for handling the uncertainties explored in SSGs; 2) based on this unified framework, we propose the first set of “unified” robust algorithms to address combinations of these uncertainties; 3) we introduce approximate scalable robust algorithms for handling these uncertainties that leverage insights across compartments; 4) we present experiments demonstrating solution quality and runtime advantages of our algorithms.
Host: CS PHD Committee
Location: 101
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Amir Houmansadr (UTexas - Austin) - The Cyberspace Battle for Information: Combating Internet Censorship
Tue, Apr 22, 2014 @ 04:00 PM - 05:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Amir Houmansadr, UTexas - Austin
Talk Title: The Cyberspace Battle for Information: Combating Internet Censorship
Series: CS Colloquium
Abstract: The lecture will be available to stream from your browser here.
The Internet has become ubiquitous, bringing many benefits to people across the globe. Unfortunately, Internet users face threats to their security and privacy: repressive regimes deprive them of freedom of speech and open access to information, governments and corporations monitor their online behavior, advertisers collect and sell their private data, and cybercriminals hurt them financially through security breaches.
My research aims to make Internet communications more secure and privacy-preserving. In this talk, I will focus on the design, implementation, and analysis of tools that help users bypass Internet censorship. I will discuss the major challenges in building robust censorship circumvention tools, introduce two novel classes of systems that we have developed to overcome these challenges, and conclude with several directions for future research.
Biography: Amir Houmansadr is a postdoctoral scholar at the University of Texas at Austin. He received his Ph.D. from the University of Illinois at Urbana-Champaign in August 2012. Amir’s research revolves around various network security and privacy problems, including Internet censorship circumvention, network traffic analysis, and anonymous communications. He has received several awards for his research, including the Best Practical Paper Award at the IEEE Symposium on Security & Privacy (Oakland) 2013.
Host: Ethan Katz-Bassett
Location: Henry Salvatori Computer Science Center (SAL) - 101
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Amir Houmansadr (UTexas - Austin) - The Cyberspace Battle for Information: Combating Internet Censorship
Tue, Apr 22, 2014 @ 04:00 PM - 05:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Amir Houmansadr, UTexas - Austin
Talk Title: The Cyberspace Battle for Information: Combating Internet Censorship
Series: CS Colloquium
Abstract: The lecture is available to stream from your browser here starting at 4 PM. Please right-click the link to open in a new tab or window for best performance.
The Internet has become ubiquitous, bringing many benefits to people across the globe. Unfortunately, Internet users face threats to their security and privacy: repressive regimes deprive them of freedom of speech and open access to information, governments and corporations monitor their online behavior, advertisers collect and sell their private data, and cybercriminals hurt them financially through security breaches.
My research aims to make Internet communications more secure and privacy-preserving. In this talk, I will focus on the design, implementation, and analysis of tools that help users bypass Internet censorship. I will discuss the major challenges in building robust censorship circumvention tools, introduce two novel classes of systems that we have developed to overcome these challenges, and conclude with several directions for future research.
Biography: Amir Houmansadr is a postdoctoral scholar at the University of Texas at Austin. He received his Ph.D. from the University of Illinois at Urbana-Champaign in August 2012. Amir’s research revolves around various network security and privacy problems, including Internet censorship circumvention, network traffic analysis, and anonymous communications. He has received several awards for his research, including the Best Practical Paper Award at the IEEE Symposium on Security & Privacy (Oakland) 2013.
Host: Ethan Katz-Bassett
Location: Henry Salvatori Computer Science Center (SAL) - 101
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS RASC Seminar: Steven M. LaValle (Oculus VR) - Virtual Reality, Really!
Thu, Apr 24, 2014 @ 11:00 AM - 12:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Steven M. LaValle, Oculus VR
Talk Title: Virtual Reality, Really!
Series: RASC Seminar Series
Abstract: It has been an exciting adventure as we race to bring the consumer version of the Oculus Rift VR headset into widespread use for games, cinema, therapy, virtual travel, and beyond. Palmer Luckey's 2012 prototype demonstrated that smartphone-based advances in display and sensing technology enable a lightweight, high field-of-view VR experience that is affordable by the masses. This has stimulated widespread interest across many industries, research labs, and potential end users of this technology. This talk will highlight some of the ongoing technical challenges, including game development, user interfaces, perceptual psychology, and accurate head tracking. Although VR has been researched for decades, many new challenges arise because of the ever changing technology and the rising demand for new kinds of VR content.
Biography: Steven M. LaValle is Principal Scientist at Oculus VR, Inc. He is a roboticist and a Professor of Computer Science at the University of Illinois, Urbana-Champaign. He is best known for his introduction of rapidly exploring random tree (RRT) algorithms, and his book on Planning Algorithms, one of the most highly cited texts in the field. In 2012, he was one of seven faculty named as a University Scholar at UIUC for 2012-2014.
Host: Nora Ayanian
Location: Kaprielian Hall (KAP) - 156
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Ariel Feldman (University of Pennsylvania) - Designing Systems for Skeptics
Thu, Apr 24, 2014 @ 04:00 PM - 05:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Ariel Feldman, University of Pennsylvania
Talk Title: Designing Systems for Skeptics
Series: CS Colloquium
Abstract:
In modern distributed systems, users are increasingly being asked to rely on third parties who do not necessarily have their best interests in mind. For example, cloud hosted services offer a myriad of benefits, but they require users to trust the service provider with the confidentiality and integrity of their data and the correctness of the computations performed on it. The recent history of accidental and malicious data disclosures, misuse of users’ data, surreptitious censorship, and warrantless surveillance has shown that this trust is often misplaced. Moreover, non-technical mechanisms, such as laws and market incentives, have proved to be insufficient to mitigate these threats.
In this talk, I will present two implemented systems that enable their users to benefit from cloud deployment, but that are designed “for skeptics:” they provide users with guarantees that hold even if the service provider cannot be trusted. The first system, SPORC, makes it possible to build low-latency collaborative Web applications such as shared text editors, group calendars, and instant messaging applications with an untrusted provider. The provider only sees encrypted data and cannot deviate from correct execution without detection. And if the provider does misbehave, SPORC gives users a means to recover. Pantry, the second system, enables a user to outsource a general purpose computation to a potentially faulty provider and yet verify that the computation was performed correctly. Unlike prior efforts, Pantry allows verifiable computations to operate on remotely-stored data that the user does not possess, opening the way to a wide variety of uses such as MapReduce jobs and database queries.
Biography: Ari Feldman is a postdoctoral researcher at the University of Pennsylvania whose research focuses on building systems that provide confidentiality, integrity, and correctness by design rather than solely through non-technical means, drawing on techniques from distributed systems, applied cryptography, and theory. He received his Ph.D. in computer science from Princeton University in 2012 under the supervision of Edward W. Felten and received an A.B. in computer science and in ethics and political philosophy from Brown University.
Host: Minlan Yu
Location: Henry Salvatori Computer Science Center (SAL) - 101
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS RASC Seminar: Mac Schwager (Boston University) - Controlling Groups of Robots with Unreliable Relative Sensing
Fri, Apr 25, 2014 @ 02:30 PM - 04:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Mac Schwager, Boston University
Talk Title: Controlling Groups of Robots with Unreliable Relative Sensing
Series: RASC Seminar Series
Abstract: Groups of robots working collaboratively have the potential to change the way we sense and interact with our environment at large scales. However, in order to be useful in the real world, multi-robot systems must perform without global information, and they must adapt to faulty sensors. This talk will describe our recent work in controlling groups of robots with unreliable relative sensing measurements. We will treat two basic multi-robot problems: formation control and coverage control. In the first problem, we would like the robots to converge to a desired formation without a shared global reference frame, using only relative distance and bearing measurements. We propose a novel nonlinear control architecture that ensures asymptotic convergence to the desired formation. We also implement this controller on a network of quadrotor aerial robots. The robots use onboard vision, computing relative pose estimates from shared features in their images, in order to execute the formation controller without any global pose information. In the second problem we consider deploying a group of sensing robots to cover an environment with their sensors, however some (a priori unknown) robots have faulty sensors. We propose a decentralized adaptive control approach by which the robots collaboratively determine which robots have faulty sensors, and reposition themselves in order to compensate for the sensor faults. Convergence to a locally optimal sensing configuration is proven using a Lyapunov analysis.
Biography: Mac Schwager is an assistant professor in the Department of Mechanical Engineering and the Division of Systems Engineering at Boston University. He obtained his BS degree in 2000 from Stanford University, his MS degree from MIT in 2005, and his PhD degree from MIT in 2009. He was a postdoctoral researcher in the GRASP lab at the University of Pennsylvania from 2010 to 2012. His research interests are in distributed algorithms for control, perception, and learning in groups of robots and animals. He received the NSF CAREER award in 2014.
Host: Nora Ayanian
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Justin Solomon (Stanford) - Embracing Uncertainty in Geometric Data Analysis
Tue, Apr 29, 2014 @ 04:00 AM - 05:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Justin Solomon, Stanford University
Talk Title: Embracing Uncertainty in Geometric Data Analysis
Series: CS Colloquium
Abstract: Many methods dealing with data on geometric domains suffer from noise, nonconvexity, and other challenges because they are forced to make choices among nearly-indistinguishable possibilities. For instance, edge-preserving image filters must assign pixels near the boundary of an object to either its interior or its exterior, inheriting different colors, textures, and other properties depending on the particular outcome. In geometry processing, algorithms for registering scans of three-dimensional objects must break discrete (e.g. left-right) and continuous (e.g. cylindrical or translational) symmetries to settle on a single correspondence.
In this talk, I will present techniques for explicitly acknowledging these and other ambiguities within graphics, imaging, and data processing pipelines. In particular, rather than making arbitrary tie-breaking decisions, these methods maintain distributions over potential outcomes. This “soft” probabilistic framework explicitly acknowledges challenging ambiguities and can be used to design robust techniques for processing and understanding images and shapes. In addition to introducing the relevant theory, I will show how it can be used to derive practical algorithms for photo processing, network analysis, and surface mapping.
Biography: Justin Solomon is a PhD candidate in the Geometric Computing Group at Stanford University. He studies problems in graphics, learning, and imaging combining techniques from mathematical theory and computer science. His work has led to practical applications in geometry processing, computational photography, and medical imaging and is supported by the Hertz Foundation Fellowship, the NSF Graduate Research Fellowship, and the NDSEG Fellowship.
Justin holds bachelors degrees in mathematics and computer science and a masters degree in computer science from Stanford. He is a dedicated instructor and has served as the lecturer for courses in graphics, differential geometry, and numerical methods. His forthcoming textbook entitled Numerical Algorithms focuses on applications of numerical methods to graphics, learning, and vision. Before beginning his graduate studies, Justin was a member of Pixar's Tools Research group. Outside the lab, he is a pianist, cellist, and amateur musicologist with award-winning research on early recordings of the Elgar Cello Concerto.
Host: Fei Sha
Location: Henry Salvatori Computer Science Center (SAL) - 101
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
