In this speech, we present UQLIPS - a real-time near-duplicate video retrieval and monitoring system. As we review state-of-the-art practices, we first investigate existing variants of the definition of near-duplicate video, discuss the main challenges, and describe a generic framework. Then we focus on the presentation of UQLIPS, which employs several novel techniques in video signature generation, indexing and query processing to achieve real-time and accurate search and monitoring over large-scale video data. Finally, we explore the emerging trends of this research topic.
18-Jan-2011: Assoc. Professor Heng Tao Shen
University of Queensland (http://www.itee.uq.edu.au/~shenht/)
UQLIPS: A Real-time Near-duplicate Video Retrieval and Monitoring System
Recorded Keynote: Play the keynote
The exponential growth of online videos, along with the increasing user involvements to video-related activities, has been observed as a constant phenomenon during last decade. User's time spent on video capturing, editing, uploading, searching and viewing has boosted to an unprecedented level. The massive publishing and sharing of videos has given rise to the existence of a already-large amount of near-duplicate content. This imposes urgent demands on near-duplicate video retrieval and monitoring as key roles in novel tasks such as video search, copyright protection, video recommendation, content tracking, redundancy removal, and many more. Driven by the strong significance, near-duplicate video retrieval and detection have recently attracted lots of attention.
Dr Heng Tao Shen is currently an Associate Professor and Reader in School of Information Technology & Electrical Engineering, The University of Queensland. He obtained his BSc (with 1st class Honours) and PhD from Department of Computer Science, National University of Singapore in 2000 and 2004 respectively, then joined The University of Queensland as a Lecturer in June 2004 and Senior Lecturer in March 2007. His research interests include Multimedia/Mobile/Web Search, Database Management, P2P/Cloud Computing, etc. Heng Tao has published and served on program committees in most of prestigious international publication venues of interests, such as ACM SIGMOD, ACM Multimedia, VLDB, VLDB Journal, ACM TOIS, IEEE TMM, IEEE TKDE, etc. He is the recipient of Chris Wallace Award for outstanding Research Contribution in 2010 conferred by Computing Research and Education Association of Australasia.
19-Jan-2011: Professor Gene Tsudik
University of California at Irvine (http://www.ics.uci.edu/~gts/)
Usable Security: the case of user-aided pairing of wireless devices
Recorded Keynote: Play the keynote
'Secure device pairing' is the process of bootstrapping a
secure channel between two or more previously unassociated
personal devices over a (usually wireless) human-imperceptible
communication channel. Lack of prior security context and
absence of common trust infrastructure open the door for
so-called "Man-in-the-Middle" (or "Evil Twin") attacks. Mitigation
of these attacks requires user involvement in the device pairing
process. This research direction is both important and timely, since it considers
usability in one of the few security-relevant settings that involves a wide range of users.
Prior research yielded a number of interesting methods
utilizing various auxiliary human-perceptible channels, e.g., visual,
acoustic and tactile. These methods engage the user in authenticating
information exchanged over human-imperceptible channels,
thus countering attacks and forming the basis for secure device
This talk will begin by summarizing notable secure device pairing
techniques, comparing and contrasting their advantages, shortcomings
and limitations. Then, we present a comprehensive and comparative
evaluation (based on both on usability and security) of these methods.
Results identify methods best-suited for a given combination of devices and
human abilities. Next, we consider the group setting where more than two
unfamiliar devices need to be associated in order to set up a secure
communication context. We then report on a usability study that compares
several potential group pairing techniques.
We conclude with some unresolved issues and potential avenues for future
Gene Tsudik is a Professor of Computer Science at the University of California, Irvine (UCI).
He obtained his PhD in Computer Science from USC in 1991. Before coming to UCI in 2000, he was at
IBM Zurich Research Laboratory (1991-1996) and USC/ISI (1996-2000). Over the years, his research
interests included many topics in security and applied cryptography. He currently serves as Director of
Secure Computing and Networking Center (SCONCE) and Vice-Chair of the Computer Science Department.
Since 2009, he is the Editor-in-Chief of ACM Transactions on Information and Systems Security (TISSEC).
20-Jan-2011: Professor Dexter Kozen
Cornell University (http://www.cs.cornell.edu/~kozen)
Computing with Capsules
Recorded Keynote: Play the keynote
Capsules provide a clean algebraic representation of the state of a
computation in higher-order functional and imperative languages with
mutable bindings. Capsules play the same role as closures or heap- or
stack-allocated environments but are much simpler. A capsule is
essentially a syntactic representation of a finite coalgebra
representing a regular closed lambda-coterm. In this talk I will
describe a simple higher-order programming language with functional and
imperative features, including mutable bindings, based on capsules. For
computational purposes, all operations of interest are typable with
simple types, yet the system is Turing complete. Recursive functions
are represented directly as capsules, eliminating the need for unnatural
fixpoint combinators that are untypable with simple types and impose a
call-by-name discipline. I will give an operational semantics for the
language and show that it correctly captures lexical scoping without any
combinatorial constructs such as stacks, heaps, or closures.
Dexter Kozen is the Joseph Newton Pew, Jr. Professor in Engineering at
Cornell University. He received his undergraduate degree from Dartmouth
College in mathematics in 1974 and his PhD from Cornell in computer
science in 1977. After working as a member of the research staff at the
IBM Thomas J. Watson Research Center for several years, he returned to
Ithaca to join the Cornell faculty in computer science in 1985. He is a
recipient of the John G. Kemeny Prize in Computing and an IBM
Outstanding Innovation Award, and is a former Guggenheim fellow and a
fellow of the Association of Computing Machinery and American
Association for the Advancement of Science. Kozen's research interests
span a variety of topics on the boundary of computer science and
mathematics: design and analysis of algorithms, computation complexity
theory, complexity of decision problems in logic and algebra, and logics
and semantics of programming languages. He is the author of over 150
research articles and four books.
Kozen lives with his wife Frances, a lecturer in the College of Human
Ecology at Cornell. They have three sons, Alex 26, a PhD student at the
University of Maryland, Geoff 24, a law student at Georgetown, and Tim
21, a senior at Cornell. For leisure activities, he enjoys music of all
types, but especially modern rock. He can be seen occasionally playing
guitar with a local Ithaca rock band. He also enjoys sports, especially
rugby, ice hockey, and skiing.