RESEARCH KEYNOTE SERIES
Prof. Vincent W. S. Chan
(Massachusetts Institute of Technology, USA)
Throughout his career, Professor Chan has spent his research focus on communication and networks, particularly on free space and fiber optical communication and networks and satellite communications. His work has led the way to a successful laser communication demonstration in space and early deployment of WDM optical networks. His recent research emphasis is on algorithmically-optimized heterogeneous network architectures with stringent performance demands.
Title of the Talk: Research frontiers in communications and networks
Abstract: Due to the advent of disruptive new technology and driven by radically new applications different from the traditional ones, the network of the future will grow at least 3 orders in magnitude in capacity and more significantly much more agile in its response to changing demands. This talk will highlight including providing conjectures on the architecture of future networks. Emphasis of the talk will be placed on new service requirements and network architecture and algorithms.
Prof. Ken Birman
(Cornell University, USA)
Bio: Ken Birman is the N. Rama Rao Professor of Computer Science at Cornell. An ACM Fellow and the winner of the IEEE Tsutomu Kanai Award, Ken has written 3 textbooks and published more than 150 papers in prestigious journals and conferences. Software he developed operated the New York Stock Exchange for more than a decade without trading disruptions, and plays central roles in the French Air Traffic Control System and the US Navy AEGIS warship. Other technologies from his group found their way into IBM’s Websphere product, Amazon’s EC2 and S3 systems, Microsoft’s cluster management solutions, and the US Northeast bulk power grid. His Vsync system (vsync.codeplex.com) has become a widely used teaching tool for students learning to create secure, strongly consistent and scalable cloud computing solutions. Derecho is intended for demanding settings such as the smart power grid, smart highways and homes, and scalable vision systems.
Title of the Talk: Building Smart Memories and Responsive Edge Services with Derecho
Abstract: The Derecho platform was created to support a new generation of Internet-of-Things (IoT) applications with online machine-learning components. At cloud-scale, such applications require a new edge u-service ecosystem, which I like to think of a as a form of “smart memory”. I’m using this term to refer to a customizable service designed to be hosted in the cloud edge, where it would accept high-bandwidth data pipelines from sources, apply machine-learning tools to analyze and understand received content, perform initial data transformations such as image segmentation, tagging and other basic AI functions, and support ways to query the resulting knowledge base with minimal delay. Such services would also need to scale out, yet must maintain their rapid responsiveness and strong consistency.
Derecho, which is now fully implemented (github.org/Derecho-Project), leverages persistent memory and RDMA to solve this problem with exceptional performance and scalability. Derecho is also interesting from a theoretical perspective.In particular, the core protocols used implement Paxos state machine replication in a novel manner optimized for RDMA settings. These protocols have been proved correct, and are also highly efficient in terms of delay before message delivery, progress during failures and even the mapping to RDMA hardware.
Prof. Amy Greenwald
(Brown University, USA)
Bio: Dr. Amy Greenwald is Professor of Computer Science at Brown University in Providence, Rhode Island. She studies game-theoretic and economic interactions among computational agents, applied to areas like autonomous bidding in wireless spectrum auctions and ad exchanges.
During the 2018--19 academic year, she was a visiting researcher at the Artificial Intelligence Research Center at the Japanese National Institute of Advanced Industrial Science and Technology in Tokyo. In 2015, she was a visiting researcher in the
Algorithmic Economics Lab at Microsoft Research in New York City. In 2011, she was a visiting professor at the Erasmus
Research Institute of Management in Rotterdam.
In 2011, she was also named a Fulbright Scholar (though she declined the award). She was awarded a Sloan Fellowship in 2006;she was nominated for the 2002 Presidential Early Career Award for Scientists and Engineers (PECASE); and she was named one of the Computing Research Association's Digital Government Fellows in 2001. Before joining the faculty at Brown, Dr. Greenwald was employed by IBM's T.J. Watson Research Center. Her paper entitled "Shopbots and Pricebots" (joint work with Jeff Kephart) was named Best Paper at IBM Research in 2000.
Title of the Talk: Learning Equilibria in Simulation-Based Games ... and the Ensuing Empirical Design of Mechanisms
Abstract: We describe a methodology for the design of parametric mechanisms, which are multiagent systems inhabited by strategic agents, with knobs that can be adjusted to achieve specific goals. For example, a network designer might seek a design that minimizes congestion assuming selfish agents. Our methodology applies under two key conditions: 1. the mechanisms induce games that can be simulated, but that do not afford an analytic description, 2. the agents play approximate equilibria in these simulation-based games. Under these conditions, we use the probably approximately correct learning framework to construct algorithms that learn equilibria. We show experimentally that our methodology can be used to design effective mechanisms that capture stylized, but rich multiagent systems, such as advertisement exchanges, which are not generally amenable to analytical mechanism design.
Prof. Lyle Ungar
(University of Pennsylvania, USA)
Bio: Dr. Lyle Ungar is a Professor of Computer and Information Science at the University of Pennsylvania, where he also holds appointments in multiple departments in the Schools of Business, Medicine, Arts and Sciences, and Engineering and Applied Science. Lyle received a B.S. from Stanford University and a Ph.D. from M.I.T. He has published over 250 articles, supervised two dozen Ph.D. students, and is co-inventor on ten patents. His current research focuses on developing scalable machine learning methods for data mining and text mining, including deep learning methods for natural language processing, and analysis of cell phone and social media to better understand the drivers of physical and mental well-being.
Title of the Talk: Measuring Well-Being Using Social Media
Abstract: Social media such as Twitter and Facebook provide a rich, if imperfect, portal into people's lives. We analyze tens of millions of Facebook posts and billions of tweets to study variation in language use with age, gender, personality, and mental and physical well-being. Word clouds provide insights into stress, anxiety, and depression, while correlations between language use and county-level health data suggest connections between health and happiness, including potential psychological causes of heart disease.
Prof. Stephen B. Wicker
(Cornell University, USA)
Bio: Dr. Stephen B. Wicker is a Professor of Electrical and Computer Engineering at Cornell University, and a member of the graduate fields of Computer Science, Information Science, and Applied Mathematics. He teaches and conducts research in wireless information networks, cellular networks, and digital telephony. He currently focuses on the interface between information networking technology, law, and sociology, with a particular emphasis on how design choices and regulation can affect the privacy and speech rights of users.
Professor Wicker’s most recent book, Cellular Convergence and the Death of Privacy, was published by Oxford University Press in August 2013. Professor Wicker received a Cornell College of Engineering Teaching Award in 1998, 2009, and 2013. He also received the Cornell School of Electrical and Computer Engineering Teaching Award in 2000. As of early 2019, he has supervised 44 doctoral dissertations.
Professor Wicker is also the author of Codes, Graphs, and Iterative Decoding (Kluwer, 2002), Turbo Coding(Kluwer, 1999), Error Control Systems for Digital Communication and Storage (Prentice Hall, 1995) and Reed-Solomon Codes and Their Applications (IEEE Press, 1994).Professor Wicker is the Cornell Principal Investigator for the TRUST Science and Technology Center – a National Science Foundation center dedicated to the development of technologies for securing the nation’s critical infrastructure. He is a Fellow of the IEEE.
Title of the Talk: Reading in the Panopticon: A Case Study in Wireless Surveillance
Abstract: In this talk I pursue the Panoptic metaphor by considering surveillance technologies that may be built into our eBooks. I choose the words “may be” with great care; my students and my studies of Amazon’s patents indicate the potential for extensive surveillance, but when we asked Amazon to confirm or deny their use of these technologies, we received what can best be described as a non-answer. It follows that Kindle users don’t know that the surveillance technologies that I will describe are actually in use, only that they are available for use. And that, of course, is the underlying power of the Panopticon.
Having described Amazon’s patented Kindle surveillance technology, I turn to the question of why we should care. Using case law and common sense, I suggest that anonymous reading is connected to free expression. Surveillance creates a chilling effect on one’s choice of reading material, which in turn limits what one has to contribute to the marketplace of ideas. I conclude with a brief discussion of possible policy solutions.
Prof. Julie Dorsey
(Yale University, USA)
Bio: Dr. Julie Dorsey is a professor of Computer Science at Yale University, where she teaches computer graphics. She came to Yale from MIT, where she held tenured appointments in both the Department of Electrical Engineering and Computer Science (EECS) and the School of Architecture. She received undergraduate degrees in architecture and graduate degrees in computer science from Cornell University. Her research interests include photorealistic image synthesis, material and texture models, sketch-based modeling, and creative applications of AI. Her current and recent professional activities include service as the Editor-and-Chief of ACM Transactions on Graphics (2012-15) and membership on the editorial boards of Foundations and Trends in Computer Graphics and Vision, Computers and Graphics, and IEEE Transactions on Visualization and Computer Graphics. She has received several professional awards, including MIT’s Edgerton Faculty Achievement Award, a National Science Foundation Career Award, an Alfred P. Sloan Foundation Research Fellowship, along with fellowships from the Whitney Humanities Center at Yale and the Radcliffe Institute at Harvard; she was winner of Microsoft's international Female Founders Competition. She is co-author of Digital Modeling of Material Appearance and the founder and chief scientist of Mental Canvas, a NYC-based software company that is developing a new type of interactive graphical media and a system to design this form of media.
Title of the Talk: The Future of Sketch: How Would Leonardo Draw Today?
Prof. Peter K. Allen
(Columbia University, USA)
Bio: Dr. Peter K. Allen is Professor of Computer Science at Columbia University, and Director of the Columbia Robotics Lab. He is the recipient of the CBS Foundation Fellowship, Army Research Office fellowship, the Rubinoff Award for innovative uses of computers, and the NSF PYI award. His current research interests include robotic grasping, medical robotics and Brain-Computer Interfaces for Human- Robot Interaction.
Title of the Talk: Teaching Robots to Grasp via Multi-Modal Geometric Learning
Abstract: Complex, high degree-of-freedom tasks such as grasping and manipulation are often difficult for robots to accomplish. AI and machine learning are promising technologies that can transfer skills from humans to robots. In this talk, we will describe methods to enable robots to grasp novel objects using multi- modal data and machine learning. The starting point is an architecture to enable robotic grasp planning via shape completion using a single occluded depth view of objects. Shape completion is accomplished through the use of a 3D CNN. The network is trained on our open source dataset of over 440,000 3D exemplars captured from varying viewpoints. At runtime, a single 3D depth image captured from a single point of view is fed into the CNN, which fills in the occluded regions of the scene,allowing grasps to be planned and executed on the completed object, which extends to novel objects as well. We have extended this network to incorporate both depth and tactile information. Offline, the network is provided with both simulated depth and tactile information and trained to predict the object's geometry, thus filling in regions of occlusion. At runtime, the network is provided a partial view of an object and exploratory tactile information is acquired to augment the captured depth information.We demonstrate that even small amounts of additional tactile information can be incredibly helpful in reasoning about object geometry.