ISSCS
2005 Program (Final)
Symposium at a Glance
Presentation
Information
For
each paper, the duration of the presentation will be 20 minutes, 15 minutes
for the presentation itself and 5 minutes for questions and answers. However,
there might be slight changes of the program (e.g. authors do not show
up) and the chairman of your session may allocate some additional time
for presentation.
In each session room a computer (MS Windows 2000/XP) with a CD-ROM drive,
USB ports and a beamer will be available for presentations. The presentations
should be made on MS PowerPoint or Adobe Acrobat Reader and should be
copied on the computer before the session begins.
Speakers are kindly asked to be in their session rooms at least 10 minutes
before the sessions begin in order to contact the sessions chairs and
to provide them the necessary information (name, affiliation, position,
main research topics) for making the introduction to the audience. A member
of the organizing committee will be in each session room to assist the
speakers.
Information
for Session Chairs
Chairmen are kindly
asked to be in their session rooms at least 10 minutes before the sessions
begin in order to meet the authors they are going to chair. A short presentation
of the authors will be highly appreciated.
In general, every presentation, discussions included, was planned to fit
into a 20 minutes interval. However, there might be slight changes in
the program (e.g. authors do not show up) and the chairmen are free to
allocate the time in a manner they are considering appropriate but paying
attention to end the sessions in time.
A member of the organizing committee will be in each session room for
assistance.
Special
Session: "Modern On-chip Communication Platforms"
Organizer:
Tiberiu SECELEANU, Univ. of Turku, Turku, Finland
As
System-on-Chip designs become larger and more complex, there has been
a shift of focus towards a platform-based design methodology that allows
the designer/architect to focus initially on the functions that a system
has to perform, separate from thinking about how those functions are to
be implemented. New solutions are emerging and may induce drastic changes
in the current design flow concepts, starting at the system level through
to the physical layout. Therefore, we are a long way from knowing exactly
how such platforms may effectively be designed.
This
session will concentrate on the realization of structures supporting on-chip
communication, that is, buses or network (NoC). The approach will mostly
focus (but not limit to) on the following areas:
-
High level modeling of bus / NoC based systems
- System models
- Communication protocols
- Architectural trade-offs
- Hardware/software communication abstractions
- Reconfigurability
(statically / dynamically)
- FPGAs
- Specific processors
- Performance
and synthesis issues
- New
technologies
- Integration
- Physical
level effects
- Power
consumption
- Tools
- Applications
Please
send the manuscript proposals directly to the session organizer,
Dr.
Tiberiu Seceleanu, tiberiu.seceleanu@utu.fi.
Special
Session:
"Geometrical and Statistical Models in Image/Video Processing"
Organizer:
Vasile BUZULOIU, "Politehnica" University of Bucharest, Romania
Image
processing is a wide interdisciplinary area both from the methodological
and applied point of view: large collections of data are measured, stored
and processed/ interpreted in order to extract from them as much information
as possible. Statistical models are powerful tools allowing to embed prior
knowledge and noise robustness into the processing tasks. Contributions
in all these areas (as well as any related topics) are welcomed.
Please
send the manuscript proposals directly to the session organizer,
Prof. Vasile Buzuloiu, buzuloiu@alpha.imag.pub.ro.
Special
Session: "Telemedicine
and Medical Image Processing"
Organizers:
Alexandrina ROGOZAN, INSA de Rouen, France
Constantin VERTAN, "Politehnica" University of Bucharest,
Romania
Digital
image processing is widely used in medical sciences, from clinical investigation
to patient record management. We feel that there is an important current
trend in the development of distance and distributed medical applications,
covering all aspects of the domain, from gathering data, maintaining medical
archives, assisting the diagnosis and the medical procedures. Contributions
in all these areas (as well as any related topics) are welcomed.
Please
send the manuscript proposals directly to the session organizer,
Dr. Constantin Vertan, cvertan@alpha.imag.pub.ro.
Special
Session: "Power-Aware
Design of Integrated Circuits and Systems"
Organizers:
Dr. Alberto GARCIA-ORTIZ, IBM/TU-Darmstadt, Germany
As
CMOS technology drives into the era of very deep sub-micron devices, novel
challenges and opportunities related to power efficiency of integrated
circuits appear. The high integration potential provided by state-of-the
art and upcoming technologies allows the single-chip design of increasingly
complex products. However, as those design requirements are pushing towards
higher operating frequencies, the so-called power crisis gets aggravated.
Thus, new paradigms as leakage, coupling capacitances, second-order transistor
effects, and higher design complexity, are changing the classical approaches
for power estimation and optimization.
The
scope of this session are novel power-aware design methodologies addressing
the challenges associated with the nano-scale era. It is mainly focused
(but not limited) to the following areas:
a) Power optimization:
- Leakage efficient circuits
- Low-power analog & mixed-mode circuits with scaled devices
- Power-aware DSP architectures
- Power-efficient on-chip bus-encoding schemes
- Dynamic power management at different levels of abstraction
- Power efficient operating systems & compilers
- Power-aware design methodologies & synthesis
b) Power estimation:
- Power models for nano-scaled transistors
- Leakage estimation
- Power estimation in complex digital architectures (Processors cores,
DSPs, ...)
- Power estimation for embedded systems
- Power estimation at high-levels of abstraction
Please
send the manuscript proposals directly to the session organizer,
Dr. Alberto Garcia-Ortiz, agarcia@mes.tu-darmstadt.de.
A
tutorial on "Motion Analysis and Video Understanding" given
by Dr. Alexandra Branzan Albu, Computer and Vision Systems Laboratory,
Laval University, Canada is planned for 13 July 2005, 16:00-19:00.
Tutorial
Audience:
This
tutorial is intended for people involved in signal, image and video processing
(researchers and engineers), as well as for undergraduate students possessing
basic knowledge in the field of signal processing.
Tutorial description:
A fundamental property of biological vision systems is their capacity
to sense and interpret motion. This information is crucial for controlling
the self-motion in the real world, as well as for detecting the motion
of other objects belonging to the observed scene. While motion information
is embedded in the optical flow generated by real scenes, apparent motion
can be perceived from a sequence of static images, commonly known as a
video sequence.
In computer vision, the analysis of video sequences is required in a variety
of applications, such as measuring rigid, articulated or non-rigid motion,
object tracking and real-time event detection. Moreover, the recognition
of motion from video sequences represents a central task of intelligent
computer vision-based systems. For example, perceptual interfaces require
real-time hand gestures and/or facial expressions recognition.
The main objective of this tutorial is to provide an in depth introduction
to the domain of video understanding, and to outline fundamental techniques
for motion analysis, modeling and recognition. Several applications in
surveillance, biometrics and medical imaging will be also discussed in
detail.
Tentative topics:
- Motion perception in biological systems.
- Extraction and spatiotemporal representation of motion information in
computer vision.
- Models for rigid and articulated motion. Applications in surveillance
systems.
- Deformable motion models applied in medical imaging.
- Motion recognition. Algorithms and performance evaluation.
|