February 13,1991
Mr. Hou Ching-Piao
Vice President
National Science Council
106, Hoping east Rd., sec.2,
Taipei
N/ref. CL 16/91/PM/tl
Dear Mr. Vice-President,
Please find enclosed herewith the proposals of the French side
concerning the scientific cooperation program between our two c-
ountries for the year 1991.
I hope that they will meet the agreement of the National Science
Council.
Thanking you in advance for your answer. I remain, Dear Mr. Vic-
e-President,
Yours Sincerely,
[Signed]
Pierre MALLET
Deputy Director
Proposals for Scientific Cooperation 1991
Ⅰ Short duration grants
They will concern the usual fields of cooperation: applied m-
athematics, geology, medical sciences, oceanography, unclear
sciences, biotechnologies.
a. Applied mathematics
INRIA (National Institute for Research in Computer Science
and Automatisms) will remain the scientific coordinator of
the French side (Prof. Alain Bensoussan and Olivier Piron-
neau).
One grant has been included in this field in the scientif-
ic cooperation programm between the two countries. The du-
ration, the research programm, the dates of the stay will
be conveyed later to NSC.
b. Geology
Professor J. Angelier of Paris Ⅵ University will remain
the scientific coordinator of the French side.
Mr. Horng Chrong-shern will undertake a research study for
six months beginning the 18th of February 1991 at the CNRS
"Laboratorire des Faibles Radioactivites" under the sup-
ervision of Professor Carlo Laj. This scholarship was gra-
nted under the 1990 program.
There will not be any another grant this year in this fie-
ld, most of the cooperation in geology being shifted to o-
ceanography.
c. Medical sciences
Professeur Pierre TRAN Ba Huy will remain the French coor-
dinator.
- one grant of eight months (including a two month langua-
ge training program) will be awarded.
- requirement: the candidate will be specialized in other-
rhyno-laringology and have already practiced in that fi-
eld.
- date of the stay: 1st November 1991-20 June 1992. the p-
eriod of the stay could be extended for another period
of six months (1st July-31st Dec. 1992).
- language training: the candidate will follow a five mon-
th course at the Alliance Francaise of Taipei before his
departure and a two month intensive training in France.
(1st Nov. -31st Dec. 91).
d. Nuclear Sciences
CEA (Atomic Energy Committee, Mr. Billecocq) will be the
French coordinator.
One grant will be awarded in 1991.
The duration, the research program and institution, the d-
ates of the stay, will be conveyed later to NSC.
e. Biotechnologies
CNRS (MM. Paoletti and Jordan)
will be scientific coordinators.
One bilateral seminar and joint evaluation missions will
take place in 1991 (cf.infra) which should lead to the im-
plementation of a cooperation program for the years to c-
ome.
No decision will be taken concerning the grants before kn-
owing the requirements of both parties.
f. Oceanography
IFREMER (Dr. Yves Henocque) will remain the scientific co-
ordinator.
Four grants will be given this year:
- one of 3 months in the field of oceano-chemistry (Labor-
atory of CNRS, Bordeaux);
- One of 5 months in the field of sedimentology (Laborato-
ire of CNRS, Perpignan);
- Two of 3 months in the field of oceano-physics (Laborat-
oire of CNRS, Toulouse and Paris).
The selected candidates must be approved by the Chinese c-
oordinators of the cooperation program in oceanography who
are:
- Prof. K.K. Liu From Academic Sinica for the first two g-
rants;
- Prof. C.T. Liu from National Taiwan University for the
two grants in oceano-physics.
They will not undertake any language training, the durati-
on of their stay in France being too short and the French
scientists who will welcome them being able to work in En-
glish.
The exact topic of their research and the dates of their
stay in France will be communicated later to NSC.
General conditions:
- During their stay in France the candidates will receive
a monthly allowance of 5000 FF (apart from those in oce-
anography who will receive a monthly allowance of 8000
FF.)
- They will get a fully comprehensive insurance (securite
sociale);
- They must be under the age of 35 (except in the oceanog-
raphy field).
Ⅱ Seminars
Three bilateral seminars will be organized in 1991.
a. One in biotechnology on "Virology and Endocrinology of Fi-
sh and their applications to aquaculture" wil be held in
Taipei (late April) in collaboration withe Academia Sinica
and universities, with the participation of at least 7 Fr-
ench scientists (CNRS,IFREMER, others institutions).
b. One in mathematics organized by Academia Sinica which will
be held in Taipei (14-18 October) on "Combinatorics and C-
omputing" with the collaboration of the CNRS and the univ-
ersities Paris Ⅶ and Paris-Sud. Six French mathematicians
will attend this seminar.
c. One in mathematics organized in Paris by INRIA. The dates,
the topic and the participants have not yet been fixed. T-
hese informations will be conveyed later to CNS.
According to our general understanding, the organization of
the seminars will be the responsibility of the inviting side
and the travel expenses at the charge of the invited scienti-
sts.
Ⅲ Missions
Seven French scientists will come to Taipei at the invitation
of NSC for a period of one to two weeks:
- four in biotechnology (evaluation mission);
- one in applied mathematics:
- one in geology;
- one in medicine.
The name, the dates of arrival and the duration of stay of t-
hese scientists will be communicated later to NSC.
Ⅳ Invitation
The French side will invite to France, for one week, M.Hou C-
hing-Piao, Vice-President of the National Science Council.
Scientific Program 1991
SUMMARY
1 Short duration scholarships: 7
- Applied mathematics: 1 (INRIA)
- Medical sciences: 1 (Hopi-al
Lariboisiere)
- Oceanography: 4 (CNRS,
IFREMER)
- Nuclear sciences: 1 (CEA)
2 Seminars: 3
- Mathematics: 2 (one in Taipei, one in Paris)
- biotechnologies: 1 (in Taipei)
3 Missions: 7
- Biotechnologies : 4
- Mathematics : 1
- Geology : 1
- Medecine : 1
4 Invitation: 1
***************************************
March 12,1991
Mrs. Cynthia T. Lee
National Science Council
N/ref. CL 29/91/PM/tl
Dear Ms. Lee,
Referring to my letter n°16/91 dated February 13, please find
hereunder the complementary information concerning the grants of
the Sino-French scientific program for 1991.
1 Applied mathematics. Short duration grant
The candidate must have the equivalent of a Master of Sciences
in Applied Mathematics and Computer Sciences. He must know C,
UNIX and have some knowledge of COMPUTER VISION.
His field of research will be: "Adaptive analysis of a stereo
pair from a priori given 3-D geometric model of a scene", (a
more detailed note on the research subject is joined herewith)
.
His field of research will be: "Adaptive analysis of a stereo
pair from a priori given 3-D geometric model of a scene", (a
more detailed note on the research subject is joined herewith)
.
This training will last eight months at INRIA, starting from
the 1st of November 1991, under the supervision of professor
Andre Gagalowicz. The knowledge of French language is not com-
pulsory.
2 Nuclear Sciences. Short duration grant
The applicant must have a BA or MA in nuclear engineering and
several years of experience in the unclear fuels field.
The scope of the training in France will be:
- to familiarize the candidate with the French R & D in fuel
design, manufacturing and testing;
- undertake research in two related fields:
‧qualification of a new fuel assembly in his own country,
either at the level of the utility, and/or at the level of
the regulatory atomy body;
‧methodology for the establishment of a fuel fabrication f-
acility in his own country.
French language is not a prerequisite. The training session w-
ill take place in concerned labs and industrial firms in Engl-
ish Language. Nevertheless a minimum amount of French language
knowledge should be acquired before departure in order to make
daily life in France more confortable.
The training period will last eight months, starting from the
1st of November 1991. If required by NSC, an intensive French
language course would take place in France during the first t-
wo months (1st Nov. -31st Dec. 1991).
I remain, dear Ms.Lee
Yours sincerely,
[Signed]
Pierre MALLET
Deputy Director
RESEARCE SUBJECT
TITLE: Adaptive analysis of a stereo pair from a priori given
3-D geometric model of a scene.
This subject is tightly related to the main research of SYNTIM
project directed by A. Cagalowicz. This project consists of t-
he creation of a feedback loop between computer vision and co-
mputer graphics through generic application which is the synt-
hesis of a "real" 3-D office. We propose the following proced-
ure to fullfil this application we analysis a 3D office using
two cameras on an horizontal plane in a stereo position, that
could be mounted on a robot head. The analysis phase has to p-
roduce a synthetiable model of the scene. Such a model consis-
ts of two parts:
- a geometric part describing all the objects of the scene us-
ing a polyhedral approximation, but also the location and e-
xtent of light sources (described using planar facets also!)
.
- a photometric part describing the reflectance properties of
the objects (faces) but also the emittance properties of the
light sources.
It is then possible, to develop a synthesis algorithm, using
the obtained model as input, in order to produce an image (or
astereo pair) of the scene related to the same scene/camera p-
osition. The idea is to compare the synthesized image(s) with
the real image output of the cameras, visually and numerically
in order to improve the model which in its turn requires the
adaptation of the analysis and synthesis techniques to minimi-
ze the crror between the two types of images.
The crucial part of this approach is the production of synthe-
tisable models. After three years of research in our laborato-
ry, we partially solved the problem of the computation of the
geometrical part of the model. We now start the analysis of t-
he photometrical part of the model.
The solution of the geometrical interpretation is obtained th-
rough the use of a world model that we compare to the computed
geometry of the scene. This comparison provides a displacement
between the world model coordinate system and the scene coord-
inate system which minimizes the mismatch between faces of bo-
th models.
SUBJECT:
Starting from the measure of this displacement, we want now to
develop a supervised and adaptive analysis technique, the con-
trol of the analysis task being performed through the use of
the world model and the world-to-scene displacement which inf-
ers the geometrical interpretation of the scene observed by t-
he two cameras. Model controlled segmentation, region matching
and stereo algorithms have to be produced. As the displacement
meansured gives an idea of the projection of the various faces
of the objects seen by the cameras on the two image planes, we
can search, based on the two original images, homogeneous reg-
ions on the vicinity of what the world model proposes and then
enhance the original segmentations, regions matchings and 3-D
facet reconstructions (stereo algorithm) obtained by the init-
ial approach.