Implementation Challenges for Advanced Multimedia Compression
Peter Pirsch,
Leibniz University of Hannover
Internet Video: The Next Big Thing
Hui Zhang, Carnegie Mellon University, USA
Tao Yang, Ask.com, USA
KEYNOTE 1
When: 09:00-10:10, Tuesday, July 3, 2007
Title:
Computer Vision in Medicine and Neuroscience: Image Guided Neurosurgery and
Computational Neuroanatomy
Speaker: Eric Grimson
Massachusetts Institute of
Technology
Abstract:
Algorithmic methods from computer vision and machine learning are dramatically changing the practice of health care and the exploration of fundamental issues in neuroscience. By coupling knowledge of tissue response, atlases of normal anatomy, and statistical models of shape variation, these methods are used to build detailed, patient-specific reconstructions of neuroanatomical structure from MRI imagery. Such structural models can be automatically augmented with information about function (using fMRI), and about connectivity (using DT-MRI) to create detailed models of a patient’s brain. These models are routinely used for surgical planning – how to reach the target tumor with minimal damage to nearby critical structures; and for surgical navigation – guiding the surgeon to the target site rapidly and safely.
By combining with statistical models of population variation, these methods can also be used to investigate basic neuroscience questions – how different are the shapes of subcortical structures between normal subjects and patients with a specific disease (such as schizophrenia or Alzheimer’s); how do these shapes change with development in children, or with administration of pharmaceuticals; how do physiological properties differ between populations (such as the local structure of fiber orientation in white matter tracts). These computational methods provide a toolkit for exploring the structure and connectivity of neuroanatomical structures, in normal subjects and in diseased patients.
Biography:
Eric Grimson is the Bernard Gordon Professor of Medical Engineering in the Department of Electrical Engineering and Computer Science, at the Massachusetts Institute of Technology. He is also the Head of the EECS Department, and is a Lecturer on Radiology at Harvard Medical School. Professor Grimson received his B.Sc. from University of Regina and his Ph.D. from MIT. He has over thirty years experience as a researcher in computer vision and medical image analysis, having published two books, over 170 refereed articles, and six patents in this area. In recognition of these contributions, he was elected a Fellow of the American Association of Artificial Intelligence, and a Fellow of the IEEE. He has also won the Bose Award for Excellence in Teaching at MIT.
Professor Grimson’s research activities in computer
vision have included stereovision, shape reconstruction, object recognition,
image databases, and activity detection and recognition. For the past twelve
years he has also been very active in the field of medical image analysis,
especially in image guided surgery, multimodal registration, and segmentation of
medical imagery. He has supervised more than 30 doctoral theses in these areas.
Professor Grimson has served as General Chair or Program Chair for numerous
international conferences including ICCV, CVPR, and MICCAI, and has participated
on numerous government panels for DARPA, NSF, and NIH.
When: 10:40-11:50, Tuesday, July 3, 2007
Title: Implementation Challenges for Advanced Multimedia Compression
Speaker: Peter Pirsch
Abstract:
Over the last two decades compression of video and
audio data become an essential part in communication, storage and presentation
of audiovisual information. Important for the introduction of new audiovisual
services is the world wide standardization. For this reason ISO and ITU
established standardization bodies. Since almost 15 years several standards
(MPEG-i, H.26x) are introduced. Because of the large variety of applications and
the need of high compression rates several profiles, levels and a large
compression tool set has been introduced. After almost 15 years standardization
activities, new compression schemes are still introduced and a sophistication of
the compression algorithms could be observed.
A prerequisite for the introduction of new audiovisual services is
implementation cost. The advances in semiconductor technology support economic
implementation. The performance requirements of video signal compression result
in a need of signal processing architectures adapted to signal processing
schemes and supporting parallel processing. The large variety of compression
algorithms calls for software implementation. As a consequence specific devoted
programmable video signal processors have been developed. Implementation
challenges result from the request on scalable architectures with multi standard
abilities and support of different resolutions from mobile communications over
TV up to HDTV. The most recent video platforms are system-on-chip (SoC)
implementations consisting of high performance video signal processors, general
purpose RISC processors, special accelerators and peripherals.
This paper will present a brief history of the development of video signal
processors and the actual status of video platforms. VLSI architectures of
state-of-the-art programmable video signal processors are presented.
Architectural means such as SIMD (subword parallelism), VLIW, multithreading and
specialized instructions will be explained. Design examples such as Texas
Instruments DaVinci platform, Toshiba Media embedded Processor (MeP), tensilica
high definition processor and the videantis v-MP2 platform are included. The
challenges to cope with the enormous range of performance and power combinations
and the software implementations on heterogeneous multi core systems is treated.
Also the impact of advanced video compression algorithms on the architectural
requirements and the future prospects of programmable vs. reconfigurable systems
will be discussed.
Biography:
Dr. Peter Pirsch received the Ing.
grad. degree from the engineering college in Hannover, Germany, in 1966, and the
Dipl.-Ing. and Dr.-Ing. degrees from the University of Hannover, in 1973 and
1979, respectively, all in electrical engineering.
From 1966 to 1973 he was employed by Telefunken, Hannover, working in the
Television Department. He became a Research Assistant at the Department of
Electrical Engineering, University of Hannover, in 1973, a Senior Engineer in
1978. During 1979 to 1981 he was on leave, working in the Visual Communications
Research Department, Bell Laboratories, Holmdel, NJ. During 1983 to 1986 he was
Department Head for Digital Signal Processing at the SEL Research Center,
Stuttgart, Germany. Since 1987 he is Professor in the Department of Electrical
and Computer Engineering at the University of Hannover. He served as Vice
President Research of the University of Hannover from 1998 to 2002. Since 2005
he is the dean of the new formed department of electrical engineering and
computer science.
His present research includes VLSI architectures and VLSI implementations for
image processing applications, rapid prototyping and design automation for DSP
applications. He provided major contributions to VLSI architectures of
programmable video signal processors for real time implementations of video
compression schemes. He is the author or coauthor of more than 200 technical
papers. He has edited a book on VLSI Implementations for Image Communications
(Elsevier 1993) and is author of the book Architectures for Digital Signal
Processing (John Wiley 1998).
Dr. Pirsch is a member of the IEEE, the German Institute of Information
Technology Engineers (ITG) and the German Association of Engineers (VDI). He was
recipient of several awards: the NTG paper price award (1982), IEEE Fellow
(1997), IEEE Circuits and Systems Golden Jubilee Medal (1999). He was member or
chair of several technical program committees of international conferences and
organizer of special sessions and preconference courses. He has held several
administrative and technical positions with the IEEE Circuits and Systems
Society and other professional organizations. Dr. Pirsch served as Vice
President Publications of the IEEE Circuits and Systems Society from 2003 to
2004. He was chairman of the Accreditation Commission for Engineering and
Informatics of the German Accreditation Agency ASIIN. Dr. Pirsch is chair of the
VDI committee on Engineering Education.
When: 08:30-09:40, Wednesday, July 4, 2007
Title: Internet Video: The Next Big Thing
Speaker: Hui Zhang
Rinera Networs
Carnegie Mellon University, USA
Abstract:
After the long anticipation by academia and industry, the age of Internet video has finally arrived. In the very near future, video will become the dominant traffic type over the Internet. Like previous generations of Internet applications such as World Wide Web that make the consumption, production, and distribution of text/image content accessible to any users, Internet video applications will not only allow any user to access any video at any time from any where, but also enable any user to produce and distribute video at any time to any where, all over the best-effort Internet. This vision of ubiquitous Internet video differs significantly from that of the 500 TV channels over a private IP network.
A variety of technologies have been developed: Content Distribution Networks and Peer-to-Peer systems have been deployed to enhance the reach and scalability of video distribution; advanced source and channel coding techniques such as multiple description codes and network coding have been devised to address the heterogeneity and best-effort aspects of the Internet. In this talk, I will discuss why Internet video will the next big thing for both industry and academic researchers. I will review the state-of-art of Internet video technologies, discuss the key challenges, and outline the potential applications and research directions.
Biography:
Hui Zhang is founder and president of
Rinera Networks, Inc., and professor in the School of Computer Science at
Carnegie Mellon University. He has done research on clean slate Internet
architecture, Internet QoS, multicast, and peer-to-peer video streaming systems.
Algorithms and software packages resulted from his research have been widely
adopted by industry and academic institutions.
Zhang was the recipient of the National Science Foundation CAREER Award in 1996
and the Alfred Sloan Fellowship in 2000. He held the CMU SCS Finmeccanica Junior
Faculty Chair from 1998 to 2002. He was elected to be an ACM Fellow in 2006. He
was the Chief Technical Officer of Turin Networks in 2000-2003.
When: 08:30-09:40, Thursday, July 5, 2007
Where: TBD
Title: Large-Scale Internet Search
Speaker: Tao Yang
Ask.com, USA
Abstract:
The Internet is having an electrifying effect on the way we live and search engines become a central mechanism to locate right information from a massive amount of text and multi-media documents on the Internet.
Ask.com has been developing a comprehensive suite of search and question-answering technology and differentiated products to help users to find what they are looking for faster. Ask.com's ExpertRank algorithm provides relevant search results by identifying topics and the most authoritative sites on the Web through query-specific clustering and expert analysis. This talk gives an overview of Ask.com's search engine, describes many of challenges faced in seeking relevant answers from billions of documents for tens of millions of users everyday, and presents some of our approaches in topic clustering, ranking, and infrastructure support for high scalability and availability.
Biography:
Tao Yang is Chief Scientist and Senior
Vice President at Ask.com. He was chief scientist and vice president of search
development at Ask.com in charge of web search division, and the chief scientist
and vice president of research and development during Teoma's startup stage.
At Ask.com and Teoma, Yang has been directly responsible for scaling search
architectures and algorithms to handle billions of documents in terms of
relevancy, freshness, and high throughput with low latency. Yang is a
co-inventor of the ExpertRank (formally Teoma) search algorithm.
Yang has also been a full professor of computer science at University of
California at Santa Barbara with over 90 technical publications on parallel and
distributed systems, Internet search, cluster-based services, and high
performance scientific computing. Yang received Ph.D. in computer science from
Rutgers University in 1993 and B.S. in computer science from Zhejiang
University, China in 1984