Lapas attēli
PDF
ePub

KOREA ADVANCED INSTITUTE OF SCIENCE & TECHNOLOGY (KAIST)

Visual Information Processing Lab, Center for Artificial Intelligence Research

[blocks in formation]

2.

Knowledge-based image understanding system for 3D scene interpretation

3. CAD-based 3D object representation and recognition

C. Facilities (PIPE 1/300 Image Processing Super Computer, Mobile robot, SUN workstations, IBM PC 386, CCD video cameras, 400-dpi image scanner, X-terminals, VGA monitors, 21-inch high resolution color monitor)

D. Researchers - Prof. Hiyun S. Yang, 2 KAIST faculty, 7 other university faculty, 12 graduate students

Image & Communication Lab

A. Research Areas

1. Image Coding and Transmission

2. Color Image Processing

3. Image Understanding and Applications

4. Dynamic Scene Analysis

5. Channel Coding and Wideband Systems

6. Character Recognition

7. Hierarchical Coding

8. 3D Image Processing

9. Texture Image Processing

10. Shape Recognition Algorithms

B. Facilities (Micro-VAX 11, SUN workstation, Gould IP8400 and Real-time video disk, KAIST vision system, color camera for TV broadcasting)

C. Researchers - Prof. Jae Kyoon Kim, Prof. Seong Dae Kim, -30 graduate students

POHANG INSTITUTE OF SCIENCE & TECHNOLOGY (POSTECH) (COMPUTER VISION
GROUP)

A. Research Areas

1. Image Processing Algorithm Developments for Parallel Processing Computer (POPA)

2. Pattern Classification Using Computer Vision

3. 3D Object Modeling

4. Stereo Vision

5. Korean Character Recognition Using Neural Networks

6. Image Understanding

7. Robot Vision

8. Medical Imaging

B. Projects (partial list)

1. On-line detection of surface faults on metal plates

2. Autonomous land vehicle navigation

3. Development of slab number recognition system by using an image processor

4. Development of range image sensor

5.

6.

Development of high-performance parallel computer for image processing
Wavelet transforms for image processing and image understanding

7. Development of robot vision system for automatic assembly of automobile parts

C. Facilities [POPA (Pohang Parallel Architecture, transputer-based); PIPE model 1 system (image processing and understanding system); Gould IP 9516 image processor (micro-VAX host); ITI 151 image processing system;

Solbourne 5/602; Sun workstations; Neuralworks; ANZAplus neurocomputer; CCD cameras; color image scanner; HP plotter; color monitors; white light projection system; PCs; PUMA-560 robot; NOVA-IO robot]

D. Faculty - Group Leader: Prof. Chung Nim Lee

KOREA INSTITUTE OF SCIENCE & TECHNOLOGY (KIST), SYSTEMS ENGINEERING
RESEARCH INSTITUTE (SERI), COMPUTER VISION GROUP

A. Research Areas

1. Image Processing System and Applications

Satellite data processing

Computer inspection

Medical imaging

CAD/CAM and graphics

Automatic character recognition

Fingerprint recognition

2. Research on Basic Software for Remote Sensing Utilizing Image Processing Systems

B. Projects (partial list)

1. Image processing system with microcomputer

2. Development of vectorized image processing software for the processing of remotely sensed data

3. Development of automatic testing system using computer vision and AI techniques

4.

Development of weather satellite data analysis technique and workstation software for image processing

5. Development of automatic car license plate recognition system

C. Facilities (SUN workstations, IBM PC/AT based graphics systems, IBM PC-386 based graphics system, CRAY-2S/ 4-128, CYBER 960, Micro-VAX 11)

D. Researchers - 4 Ph.D., 4 M.S., 5 B.S.

FIRST NEW INFORMATION PROCESSING TECHNOLOGY WORKSHOP '91

The First New Information Processing Technology Workshop '91, held
from 5-8 November 1991 in Yokohama, Japan, is briefly described.

by David K. Kahaner

INTRODUCTION

NIPT (New Information Processing Technology) was last year's name for a major 10-year Japanese computing project proposed by the Ministry for International Trade and Industry (MITI). This year's names are Real World Computing (RWC) and FourDimensional Computing (4DC). (Japanese scientists prefer RWC, and in subsequent reports I will use this name.) The project is in feasibility study now, and if deemed "feasible" (which appears likely) it will begin officially 1 April 1991 for a period of 10 years. The suggested budget is on the order of $40-50M per year. This is a large, but by Western standards not a huge, program, although it has high visibility, and participating in it is viewed as prestigious. MITI wants to make this an international program and has offered the possibility of participation to scientists outside of Japan. The purpose of the current workshop was to listen to proposals from interested scientists. The workshop was open to those scientists whose proposals had sufficiently interested members of the steering committee and to official observers. In addition to myself, two other U.S. scientists were observers:

Dr. Marc H. Brodsky
IEEE Technology Administration
Fellow

U.S. Department of Commerce
Technology Administration

[blocks in formation]

who is experienced in the area of opto- the United Kingdom, The Netherlands,

electronics, and

Dr. Barbara L. Yoon Program Manager

and Italy, but not from the United States.
The absence of proposals from U.S.
scientists was a significant topic of
conversation during the workshop, and

Microelectronics Technology Office I was asked about it repeatedly. There
DARPA/MTO

3701 North Fairfax Drive
Arlington, VA 22203-1714
Tel: (703) 696-2234
Fax: (703) 696-2201

E-mail: byoon@a.darpa.mil

who is experienced in the area of neural nets. As much as possible, we attended different sessions. The comments below represent our joint opinions, and these were almost entirely shared by the other Western attendees. I am very grateful for the enthusiastic assistance they provided to me, and I'm also happy to report that many of my Japanese colleagues were equally willing to provide me with frank and candid opinions.

Mr. Brian Enochs, representing the Science Counsellor in the U.S. Embassy, Japan, was also present at the plenary sessions.

Approximately 80 scientists attended this workshop as shown in Table 1. There were also seven or eight MITI representatives, two Japanese observers from

was a general feeling of disappointment about the lack of collaborations with U.S. scientists, but some of the Japanese scientists felt they had plenty of their own ideas. The Europeans, while mildly concerned about the implications, were hoping that this left a larger slice of the pie for them.

The U.S. Government has suggested an independent collaboration on a rapid prototyping optical fabrication facility along the lines of the MOS Implementation Service at the University of Southern California. [A brief description of the thinking that led to this decision is given in the 28 October 1991 edition of New Technology Week (NTW).JH. Fukuda, MITI, announced it was his opinion that the United States and Japan will choose to conduct joint optoelectronics research and development (R&D) under their bilateral Science and Technology Agreement and towards that end would conduct a feasibility study.

[blocks in formation]

two-thirds, was thought to be for the
United States and one-third for the EC
and Asian countries. (Now that the
United States has declined to collabo-
rate directly, it isn't clear how this will
affect the funding.) The $400M would
also be split about equally among four
topics: (1) theory and new functions
for application, (2) massively parallel
systems, (3) neural computing systems,
and (4) optical computing systems.

There was no official European Community (EC) representation. Instead, R. Eckmiller (University of Dusseldorf, Germany) explained to the audience that the EC Commission had told him it wanted symmetric collaboration, i.e., university with university, industry with industry (if companies so desired), government institute with same, etc., and did not want oblique collaboration. It would also consider a totally vertical collaboration from basic research all the way to marketable product (that is, beyond prototype development). (Symmetric collaboration sounds great but might be difficult Functional: to achieve given the differences in research systems between Japan and the West. What about symmetric benefits?)

A very clear distinction was made between NIPT and the 5th Generation project. I quote this below.

The characteristics of the NIPT Program:

(1) To process a variety of information and to resolve such problems that are impossible or very difficult to be resolved in a logical

manner.

(2) The applicable fields are the fields

of the new information processing such as pattern information processing, resolution of the highly difficult problems, and intellectual robot, which need the functions of flexible recognition, problem solving, and control.

The characteristics of the fifth genera- System:
tion computer:

tion such as characters and to make
inference in accordance with the
logical rules, typical of the syllogism.

(2) The applicable fields are the intellec

The characteristics of the fifth gener

(1) This system is the highly parallel system composed of the element processors, and each element processor works under the control of hierarchical distributed control.

I asked about the budget for this (1) To process the symbolic informa- ation computer: program and was told that it would be of a similar order of magnitude as the ending 5th Generation project [Institute for New Generation Computer Technology (ICOT)], thus about $4050M per year. However, the exact figures are not yet available, nor are the relative percentages for university, industry, government laboratory, and foreign participation. One source said the money would be divided several ways. About 25% would be designated for foreign groups as part of the "international cooperation." Of this, $100M, about

tual information processing fields
such as the analysis of the gene
arrangement, the case study, and
judgment based on the provisions
of the laws and regulations and
judicial precedents.

(2) The parallel operating system and the application software systems are described in the logic programming language.

(3) The VLSI and ULSI technology are applied as the device technology.

« iepriekšējāTurpināt »