Re: Computer 3-goD

[P3D Gabriel Jacob <jacob@xxxxxxxxxxxxxx>]

Gregory J. Wageman writes

>Translating a single 2D image into a 3D image requires the computer to
>recognize the *content* of the image in order to determine what the
>various shapes in the image are, which are connected and which are
>overlapping, and thus decide which elements would naturally be in the
>foreground and which in the background.  This requires reasoning and
>judgement.  I'm not sure at  what age a human being acquires these
>skills, but you're talking about a computer with at least the
>cognitive abilities of an older child! 

Yes I am fully aware of these issues and that is similiar to what
I was alluding to in reference to language translation, which finally
is achievable. That to translate accurately the computer has to not
translate blindly but has to "understand" the inference and the context
of the passage. The understanding of computers in language transalation
is still nowhere near what will be capable in the future thou.

>To what program are you alluding that can do this?  I don't believe
>there is one.

There are 2 consumer ones that I am aware of and probably others that they
are working on. As I mentioned they are verrry crude to the point of being
a joke IMO. The first one is called 3D-IT! it comes on CD and is published
by Pro One Software a division of Sofsource Inc. They have a tech.support
hotline:505-523-6200 I don't know if they still exist. It was published in
1995. There address is P.O. Box 16317 Las Cruces, NM 88004.
Basically the advertising on the box says

>View All of Your Favorite Image Files in Spectacular 3D!
Includes hundeds of exciting, ready to view 3D stills and five 3D movies.
Convert and view AVI video files to 3D if you own Video for Windows.
A remarkable breakthrough in 3D digital image processing now available
at your desktop.<

Now before you all go out and buy this program I must warn you that it
is a pile of crap. They supply two algorithms that you can use and
combine but basically from what I can see it just adjusts the stereo
window!
The other program which is a European title called The Artistic Nude in 
3D. With this one you can't make your own 3D pictures but they have
nude pics in 2d and 3d format. They only have one view of the 2d pics
and it seems they use a different algorithm compared to 3D-IT. In 
both cases BTW they are presenting the 3D images as anaglyphs. The
3D anaglyph processing seems to be different in the 2nd program and
achieves the 3D effect slightly better, but not much greater.
Another don't buy product. Their email address is microapp@xxxxxxx
francenet.fr  Address available if anyone wants it.
Why do I point to these programs if they don't really work? The only
point I make is that this only the infancy and they do have along
way before 2d 2 3d pushbutton technology is feasable. And also to
point out that programs do exist, bad they may be. :-( 

>Computers also don't play chess anything at all like people do.  Most
>computer chess algorithms are brute-force implementations that rely
>on the speed and "perfect" memory of the computer to play out every
>possible variation of legal move, evaluate the resulting position,
>play out every legal answering move, etc., and "prune" this tree of
>possibilites down to the moves that give the best end-result.  People
>don't do this kind of processing.  And these computers didn't "learn"
>to play chess, they were programmed to by human intelligences!

Again I am fully aware of that aspect but what I was trying to imply here
was that initially it was thought that computer chess playing machines 
would be impossible to implement and that if they ever succeeded it indeed
would be a machine with intelligence, which of course we know is not the 
case. In any case, there are experiments carried out to so that computers
can be programmed to learn. Be it chess or whatever. The brute force method
I think would not be practical for machine vision or other more complicated
programming applications since we have reached our limits of programming
computers and will have to rely more and more on the computers programming
themselves by learning as a child does.

>Our brains are also nothing like computers.  Did you ever "forget"
>something, only to recall it a some later time?  A computer that did
>that would be considered broken.  On the other hand, stored-program
>digital computers have absolutely no imagination or intuition.  There's
>been some speculation that the human brain uses quantum phenomena at
>the lowest level, which is where our creativity and 'randomness'
>come from.  The closest thing to the human brain in computer-land is the
>neural network, and that is only a crude approximation of a few tens
>of neurons.

Yes our brains are nothing like computers but that doesn't mean we 
can't make them in the future "simulate" our pattern of thinking. 
Just like the example of playing chess where they thought you would
need a computer with INTELligence capabilities, they might be able
to replicate the thinking process by whatever means.
As you say and is what I am basically saying, we still have a long way to
go and this is why I mentioned if it will be possible in our lifetimes.
As for quantum phenomena, the way very large scale integration is going we
might reach the quantum scale of physics in that respect. Neural networks
is in the infant stage. Who knows what they will be using, maybe nano-
technology.

>And probably ant has more thinking ability than the most sophisticated
computer.

Amen. ;-)

Gabriel 




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