[comp.cog-eng] Image syntax

toyosawa@nttvdt.NTT.JP (Satoshi Toyosawa) (04/09/91)

Hi, this is my first attemp to post this newsgroup. 

In article <1991Apr7.184708.22888@colorado.edu> ralex@tigger.Colorado.EDU (Repenning Alexander) writes:

 >>Maybe part of the problem with picture-oriented representations is
 >>that, in contrast to English, picture-oriented representations do not
 >>have any defined syntax. You can of course make up a syntax using
 >>color, spatial features, explicit relationships (e.g., arcs), etc. 

That is partially  true,  and maybe too  naive, because film-makers or
artists  have some kind of  techniques  to display their  feelings and
statements  in their  pictures,   drawings, or  motion-pictures.   For
instance, there is a certain  movie tech.  to  show  a girl in  mental
shock.   Paintings,  too,  does have its  own   grammer,  like certain
shadowing on a  figure  developes  some   kind  of effect  (depresion,
darkness, etc).  Oh, BTW, honestly speaking,  I don't often understand
what a  picture trying  to  state, and don't have  skills to tell what
technique developes what emotions :-) The thing is that those who have
not  been trained to  appreciate  arts  have difficulty in  describing
pictures   in  definitive  way;  on   the  other  hand,   most of  the
arts-critics conduct it, and  that is the  reason of their existence. 
Maybe one way to introduce syntax or  grammer  of pictures is to build
an expert-system on arts. But if we  go so  much into generalizing the
essence of arts, many of good information such as indescriable feeling
and emotion will be lost, however, that maybe the good starting point,
I think.

Well, the above is my impression. Anyone who has a good taste on arts,
please make commonts.

--
----  Satoshi Toyosawa
      Visual Media Lab., NTT Human Interface Labs.
      Take 1-2356, Yokosuka, Kanagawa 238-03, Japan
      e-mail: toyosawa%nttvdt.ntt.jp@relay.cs.net

thom@garnet.berkeley.edu (Thom Gillespie) (04/09/91)

In article <1292@nttvdt.ntt.jp> toyosawa@nttvdt.NTT.JP (Satoshi Toyosawa) writes:
>Maybe one way to introduce syntax or  grammer  of pictures is to build
>an expert-system on arts. But if we  go so  much into generalizing the
>essence of arts, many of good information such as indescriable feeling
>and emotion will be lost, however, that maybe the good starting point,
>--
>----  Satoshi Toyosawa

This is an aside but Pam McCorducks new book, Aaron's Code is a must read for
anyone remotely interested in AI or expert systems and Art. It deals with the
work of harold Cohen, a painter who was world famous in the late 60's, who
decided to try to embody in art in a program. He was thedomain expert, and
became the ultimate knowledge engineer and programmer to produce Aaron, an
amazing program which draws 'like' Howard would except that there is a fair
degree of randomness built into the program. This program does not produce
'computer art' it produces art. This is agigantic step past a mere expert
system of art. The book is well written and raises all 'the ai questions' to a
new level of discussion.
\
--Thom Gillespie
.

oppizzi@butterfly.inria.fr (Olivier Oppizzi) (04/16/91)

--------------------------------------------------

I have much more taste on arts than knowing.
I have all the same worked on it, especially about "art and science
in the early of our century".
I practice a little too.

And my opinion is that art is an interpretation of the reality of both our mind and our body. Feelings flash through cognitive and sensitive filters; and afterwards
you sometimes need represent them.
I don't think emotion for instance is art.
I think art arises from mind and senses.
In what way ? I have no ansmer. Let's have a little talk anyway...

Human mental approach is conceptual and may be formalized by rules.
Or may be not.
To know, we have to think of it.
Picasso painted what he knew of a face, all sides on the same picture at the same position.
Cognitive psychologist can help to understand artists.
Semiotic for example afforded to build such rules. A result is true if it fits with enough rules. If we reverse semiotic, our thinking items become signs.

Is that much more difficult for our senses ?
A certain music, so a certain formalism, brings fear, another brings joy...
But whatever they bring, those musics mean. And once they have been detect, our
reversed semiotic is able to produce the matching sign !!!

In an effort to built sets of rules, I believe structure does matter above all.
What we should put in it depends on centuries, cultures, ourselves...
"Move" means a lot for an artist.
It must mean something for rules...





        * * * * * * * * * * * * * * * * * * * * * * *
    ********************************************* * *
                                                * * *
        Olivier OPPIZZI                         * * *
        Projet SECOIA                           * * *
        2004, route des Lucioles                * * *
        06561 Valbonne - France                 * * *
        Tel: 93.65.77.48                        * * *
        e-mail: oppizzi@mirsa.inria.fr          * * *
                                                * * *
                                                * *
    *********************************************