Copy Link
Add to Bookmark
Report
VISION-LIST Digest Volume 11 Issue 17
VISION-LIST Digest Fri May 01 14:09:18 PDT 92 Volume 11 : Issue 17
- Send submissions to Vision-List@ADS.COM
- Vision List Digest available via COMP.AI.VISION newsgroup
- If you don't have access to COMP.AI.VISION, request list
membership to Vision-List-Request@ADS.COM
- Access Vision List Archives via anonymous ftp to FTP.ADS.COM
Today's Topics:
Multiresolution motion estimation
Camera Manufacturing- Info wanted
UK MasPar Challenge
Postdoctoral Fellowship
Symposium on Physics-Based Vision at CVPR'92
Triangle NN talk: Mingolla
----------------------------------------------------------------------
Date: Wed, 29 Apr 1992 15:12:39 +0200
From: " (GIL Sylvia)" <gil@cui.unige.ch>
Organization: University of Geneva, Switzerland
Subject: Multiresolution motion estimation
Hello,
I am starting a Ph.D. program in computer vision on the subject of motion
estimation. The basic idea of the approach is based on the use of
multiresolution representations. Possibilities could be wavelets, Gabor,
Gaussian filtering.
Before starting implementing my system. I would like first to have an
idea of the existing literature on the subject. I would appreciate a lot
comments, suggestions and references.
Thanks in advance,
Sylvia Gil
Computer Vision Group, Computer Science Center
University of Geneva Tel.: +41.22.7876585
12, rue du Lac Fax: +41.22.7353905
CH - 1207 Geneva Telex: 423801 UNI CH
Switzerland E-mail: gil@cui.unige.ch
------------------------------
From: ioannisk@grip.cis.upenn.edu (Ioannis Kakadiaris)
Date: Fri, 1 May 92 15:46:34 EDT
Subject: Camera Manufacturing- Info wanted
Dear Friends:
I would to solicit information on which companies manufacture
custom made cameras. Your help will be highly appreciated.
Thank you,
Ioannis A. Kakadiaris Personal Address
University of Pennsylvania 2215 Arch Str Apt 318
School of Engineering and Applied Science Philadelphia, PA 19103
GRASP Laboratory Tel: (215) 569-1425
Department of Computer and Information Science
3401 Walnut Street, Room 311C
Philadelphia, PA 19104-6228
Tel: (215) 898-8741 (Office)
(215) 898-8361 (Lab)
Fax: (215) 573-2048
e-mail: ioannisk@grip.cis.upenn.edu
------------------------------
Date: Tue, 28 Apr 1992 14:44:48 GMT
From: jagger@maspar.com (Nigel Jagger)
Organization: MasPar Computer Corp.
Subject: UK MasPar Challenge
[ For now, this offer only applies to the UK. There are plans to
extend it to the USA. phil... ]
THE MasPar CHALLENGE
To demonstrate to everyone the ease with which massively
parallel computer systems can be programmed, MasPar is
throwing down the gauntlet to every University and Research
Council in the UK.
The challenge is to port an existing program to run on the
MasPar MP-1, in less than 3 months. The program should be
an existing, useful code but not necessarily a complete
application program.
To support this challenge MasPar will be providing a 4,096
processor MP-1 system, that will be located at the
Rutherford Appleton Laboratory (RAL) and accessible over the
JANET network. This system represents one of the smaller
MasPar configurations yet is able to deliver over 300 Mflops
and over 6,000 MIPS. The MP-1 configuration will include
the MasPar Programming Environment (MPPE), MasPar Fortran (a
version of Fortran 90), MPL (a parallel implementation of
ANSI C) and VAST-2 (which will assist the translation of F77
to F90).
To participate in the challenge, you should submit a brief
(less than 2 pages) description of your project, by
electronic mail to:
challenge@mpread.co.uk
or by regular mail to:
MasPar Computer Ltd
8 Commerce Park
Brunel Road
Theale
Berkshire RG7 4AB
Applications may be submitted now and within 5 days you will
be notified if your proposal is accepted. You will then be
required to attend a free, one day hands-on training
session. At this training you will be provided with details
of how to remotely access the MasPar MP-1 system and will be
given your personal set of MasPar documentation.
The MasPar MP-1 system will be located at RAL until the end
of August 1992, but we may be able to provide you with
access to a MasPar system beyond this date if your project
requires more time.
During the challenge period, participants will be able to
call a helpline for support and technical assistance. At
the end of the challenge the entries will be judged by a
panel of experts, based on a number of criteria which will
include:-
Performance attained
Relative speed-up
Novel implementation
Innovative algorithms
Towards the end of the year we will also be organising a
symposium where you will be able to present the results of
your brief experience with the MasPar MP-1.
We hope that you will accept our challenge based on its
technical merits alone but there will also be some prizes
for the best entries. The overall winner will be offered
the chance to spend two weeks at MasPar's headquarters in
Silicon Valley, California where they will be able to
continue their work with the support of our highly talented
engineering team. All travel and accommodation costs will
be paid by MasPar. In addition, three runners up will be
awarded #50 book tokens.
We trust that you will find this an enjoyable way to
experience the inevitable future direction for high
performance computing.
------------------------------
Date: Sat, 25 Apr 92 16:45:14 -0400
From: Jonathan Marshall <marshall@cs.unc.edu>
Subject: Postdoctoral Fellowship
A postdoctoral research position is available in the Vision Laboratory
of Christina Burbeck in the Cognitive Psychology Program, Psychology
Department, University of North Carolina at Chapel Hill. The position
is available now but could be assumed any time within the next 6
months. The position is for a 2-3 year period. Salary range is
$21-27,000 per annum.
Research in this laboratory focuses on the basic visual processes
underlying judgment of spatial relations such as the size of and
distance between objects and formation of regions into perceptual
objects. The broader theme is use of the most rigorous psychophysical
techniques to address important perceptual problems. Research in
related areas is, of course, also always of interest.
The laboratory is equipped with modern stand-alone, PC-based visual
display systems, which are supported by a half time computer
programmer. In addition, a link to the Computer Science Department
and optical disk storage in the laboratory permits use of any image
that can be generated on the massive computing facilities of the
Computer Science department.
There is active and close collaboration between members of the Vision
Laboratory and several professors and graduate students in other
departments. A group of approximately 15 professors and graduate
students from the Psychology, Computer Science, Radiology, and
Bioengineering Departments meets weekly during the academic year to
discuss vision issues. Much of the work done by these scientists is
applied in nature, with particular emphasis on medical imaging
applications and virtual environments. This research has a solid
foundation in basic research, to which the Vision Laboratory
contributes.
A strong team spirit characterizes the group. Applicants please call
or submit resume, two references, one representative research paper,
and a brief statement of research goals for the postdoctoral period
to:
Dr. Christina A. Burbeck (burbeck@uncvx1.bitnet)
Cognitive Psychology Program
Psychology Department
University of North Carolina at Chapel Hill
Chapel Hill, NC 27599-3270
(919) 962-5038.
------------------------------
Date: Tue, 28 Apr 92 21:51:48 -0400
From: Steven_Shafer@IUS5.IUS.CS.CMU.EDU
Subject: Symposium on Physics-Based Vision at CVPR'92
SYMPOSIUM ON PHYSICS-BASED VISION
Invitation to the SYMPOSIUM ON PHYSICS-BASED VISION
To be held in conjunction with CVPR-92, June 18-19, 1992
At the University of Illinois at Champaign-Urbana.
Organized by: Steve Shafer, Carnegie Mellon University, chairman
Glenn Healey, University of California at Irvine
Larry Wolff, The Johns Hopkins University
Stay one day after CVPR-92 and you can attend this unique workshop to
review the accomplishments and progress in the important area of
physics-based machine vision.
Speakers: A selected group of fourteen leading researchers will present
their views: David Forsyth, Brian Funt, Glenn Healey, Katsu
Ikeuchi, Gudrun Klinker, Shree Nayar, Alex Pentland, Steve
Shafer, Murali Subbarao, Konstantinos Tarabinis, Brian Wandell,
Reg Willson, Larry Wolff, and Bob Woodham.
Format: The speakers will give "keynote-style" presentations with their
answers to these questions: What was the state of the art in
this area before you started your work? What have you and
others accomplished? What are the "solved" problems? What
remains to be done?
Readings: There will be no proceedings or new papers. Instead, the
organizers have assembled an archival collection of 90 key
papers covering all aspects of physics-based machine vision,
which will be published in a three-volume hard-bound set by
Jones and Bartlett Publishers. This set will retail for about
$150, but at the Symposium, all attendees will receive a copy
for a deeply discounted price that is included in the
registration fee. We will ship the book set to you if you
request it (on the registration form for the symposium).
Time and Place: The Symposium will begin on the evening of Thursday,
June 18, 1992, after the last session of CVPR-92 and in the
same conference facility, with a session on Color. Starting
Friday morning, June 19, will be two sessions on Radiometry
and Camera Modeling. The Symposium will end by 2:00 pm on
Friday afternoon. Exact times are yet to be determined.
The registration fee is $160, including the readings. Hotel,
travel, etc. should be arranged in conjunction with your arrangements
for CVPR-92 itself.
For more information about the Symposium, contact Steve Shafer at
sas@cs.cmu.edu. To request travel or registration information or a
brochure for CVPR'92, call the University of Illinois at Urbana-
Champaign, Division of Conferences and Institutes, at 217-333-2883.
SPONSORED BY THE ROBOTICS INSTITUTE OF CARNEGIE MELLON UNIVERSITY
REGISTRATION FORM:
Print this, fill it out, and either FAX it to 217-333-9561
(24 hrs every day) or mail it to:
University of Illinois at Urbana-Champaign
Accounting Business Office
Room 162 Administration Building
506 S. Wright Street
Urbana IL 61801 USA
ID Information: UFAS Account # 1-3-62802-0660
REV: Computer Vision and Pattern Recognition
Last Name:
First Name and Middle Initial:
Title:
Organization:
Address for Mailing:
Telephones: Business:
Home:
FAX:
Internet E-mail Address:
Registration fee is $160, which includes a 3-volume set of
readings. Indicate one payment method; funds are payable to
"University of Illinois":
() Check enclosed (on US bank)
() Money Order enclosed (on US bank)
() Purchase Order enclosed (from US organization)
() Charge credit card: () VISA () Master Card
() American Express () Discover
Card Number:
Expiration Date:
Signature of Card Holder:
() Check here for us to ship the three volumes of readings to
your mailing address. Otherwise, you can pick them up at the
symposium itself.
------------------------------
Date: 26 Apr 92 18:46:36 GMT
From: marshall@marshall.cs.unc.edu (Jonathan Marshall)
Subject: Triangle NN talk: Mingolla
Triangle Area Neural Network Society announcement
Prof. ENNIO MINGOLLA
Cognitive and Neural Systems Department
Boston University
Thursday, April 30, 1992
5:30 p.m. Refreshments will be served.
A business meeting (elections) for TANNS will be held at 5:15.
Bell Northern Research (BNR), Research Triangle Park, NC
PROCESSING OF SYNTHETIC APERTURE RADAR IMAGES BY THE BOUNDARY CONTOUR SYSTEM
AND FEATURE CONTOUR SYSTEM
An improved Boundary Contour System (BCS) and Feature Contour System
(FCS) neural network model of preattentive vision is applied to large
images containing range data gathered by a synthetic aperture radar
(SAR) sensor. The goal of processing is to make structures such as
motor vehicles, roads, or buildings more salient and more
interpretable to human observers than they are in the original
imagery. Early processing by shunting center-surround networks
compresses signal dynamic range and performs local contrast
enhancement. Subsequent processing by filters sensitive to oriented
contrast, including short-range competition and long-range
cooperation, segments the image into regions. Finally, a diffusive
filling-in operation within the segmented regions produces coherent
visible structures. The combination of BCS and FCS helps to locate
and enhance structure over regions of many pixels, without the
resulting blur characteristic of approaches based on low spatial
frequency filtering alone.
Research performed jointly with Dan Cruthirds, Alan Gove, Stephen Grossberg,
Nick Nowak, and Jim Williamson.
For more information:
Jonathan Marshall (UNC-CH, 919-962-1887, marshall@cs.unc.edu).
------------------------------
End of VISION-LIST digest 11.17
************************