Minutes of PAMI TC Meeting June 6, 1989
San Diego, CA

Item 1.  The previous minutes were approved with an addition to
Item 8 to clairfy the context of the motion.

Item 2.  Announcements.  Tom Huang thanked Jake Aggarwal for his
service as the past chair.  The appointed officers were announced:
Treasurer: Jake Aggarwal, Secretary: Keith Price, Vice-Chair: Eric
Grimson.

Item 3.  CVPR89.  (Rama Chellappa)  330+ registered, thanks to John
Kender and Worthy Martin for the program, 327 papers submitted (90
accepted), 92 foreign (19 accepted), 22 industrial papers.  The numbers
are comparable to past meetings.  He also observed that this is the
third Computer Vision meeting in 2.5 months, all in California, and this
may reduce the attendance somewhat. 

Item 4.  Motion Workshop.  (Brian Schunk)  115 attended, 75 papers
submitted, 44 accepted, + 2 invited.  The income was about $20K, and the
expenses were about $13.4K sinced it was budgeted at 55 people.
Attendees were about evenly divided among members, students, and
outsiders (psychology mostly).

Item 5.  ICCV2  (Azriel Rosenfeld)  The newsletter is a major expense,
it now goes to the entire PAMI subscriber list.  J. Aggarwal
commented that the TC keeps a portion of any conference surplus with an
account limit of $30K at any one time.  The surplus helps ease planning
for later meetings.  ICCV-2 was close to break even or a small loss.

Item 6.  3-D Workshop Austin TX, Nov 27-29, 1989.  (Anil Jain)  The
The Call for Papers is out.  The meeting is single track, planning for
<100 people.   

Item 7.  Newsletter.  (Azriel Rosenfeld)  The mailing is to the North
Americal individual subscribers of PAMI, approximately 10K people.  The
mailing costs dominate at about $8K to $10K each.  The IAPR newsletter
is 4 per year and goes to the same list as a part of the TC newsletter.

Item 8.  ICPR.  (Jake Aggarwal)  Atlantic City, June 17-21, 1990 (new
dates).  There are 4 distinct tracks: Computer Architecture, Signal
Processing, Pattern Recognition Systems, and Computer Vision.  A lot of
papers are expected, but only 50-60 per track (240 total) will be
accepted. 
Discussion on paper topics not mentioned in the call (e.g. medical image
processing, VLSI designs) and load balancing between tracks, and
expanding to a 5th track if space is available.  Additional comments
regarding the sudden change in ICPR from most submitted papers to
highly selective papers. 

Item 9.  ICCV-3.  (Avi Kak) Will be held in Osaka, December 4-7, 1990,
at the International House (downtown).  The review procedures will be
similar to ICCV-2 (blind reviews, program committee, etc.), with the
final decisions by the program co-chairs (Tsuji, Olaf, and Kak), they
expect about 300 papers submitted with 90 accepted (30 long, 60 short). 
The fee should be about $250 with an expected attendance of 300 (200
from Japan).  The program committee totals 34 people (3 program chairs,
and 10 from Asia, 10 from Europe, and 11 from North America).

Questions regarding local expenses: under $70 for hotel is possible,
some as low as $45.  Osaka is considerably cheaper than Tokyo.  The
deadline will be known soon and some attempt is being made to raise some
travel money.  Tsuji feels that implementations (hardware and software)
should be included in ICCV-3, since they are an important part of the
work in Japan.

Item 10.  Transactionns-PAMI.  (Anil Jain for S. Tanimoto)  The
editorial board added N. Ahuja (replacing W. Thompson) and H. Baird
(replacing C. Suen).  The backlog has decreased due to the increased
page count (from 900 to 1350) and is now at about 1 issue.  Proposals
for special issues in AI, Search in Vision, probabilistic reasoning,
mathematical approaches of expert systems, etc. are desired.  The
submissions are at about 20-30 per month with about 12 papers in each
issue. 

Item 11.  No admendments to consider.

Item 12.  Discussion of standards for papers, and conferences.  (Tom
Huang) This issue grows from concerns about the different meetings
including review procedures, size (number of papers or people), format,
areas covered, etc.  A discussion of how to deal with these and how to
deal with selectivity issues followed.   Some of the highlights of the
discussion are: 
<ul>
<li> Favor wider acceptance criteria (larger number of papers) since the
acceptance rate is currently less than T-PAMI and the review is so much
more rushed. 

<li> Favor broader based conference including image processing and pattern
recognition, all in one place.  This gives a conference ranging from AI
techniques to PR techniques. 

<li> Strong desire for one strong conference with a few of the best papers.

<li> Desire for filtering truely bad papers from conferences -- too many poor
papers weakens the conference.

<li> Proposal to change ICCV to be a showcase conference that unifies the
directions by drawing on the best few papers for the other meetings
(CVPR, European CV, workshops, etc.).

<li> This committee (TC members) should set the size of CVPR (number of
tracks, number of days, length of papers) to reduce future outside
pressures on the program chairs to add papers.

<li> The current size of 300+ people and 100 papers is about right.

<li> Comments on why both ICPR and ICCV exist.

</ul>

The result of the discussion is that the executive committee of the PAMI
TC will prepare a proposal regarding conferences that addresses these
issues for the next meeting.

Motion: (T. Pavlidis) To suggest an increase in the PR and CV tracks at
ICPR to approximately 75 papers each by acquiring a fifth room.
Yes: 26, No: 0, Abstain: 2.  Passed.

Item 13.  New business.  CVPR 1991 (approximately June).  Need
proposals within 6 months or less.  A short presentation for holding the
meeting in Hawaii was well received.

A discussion regarding dues for membership in the TC.  This is now being
allowed by the Computer Society.  The question is wheather to charge
them, how much to charge, what would they provide, what services would
members want.  No decisions made.

Workshop on Robust Vision Techniques (Haralick) within 14 months.
Approved (Yes: 18, Abstain: 8).

The meeting ended with a thanks to Worthy, John and Rama for the success
of CVPR-89 and adjourned at 10:30.