From: phil osborn (philosborn@hotmail.com)
Date: Fri Jul 21 2000 - 23:52:51 MDT
In the 1986, "Mandala" appeared on the Amiga, the first "videoplace" to run
on a personal computer. When the initial review appeared in the TPUG, the
magazine of the world's largest user group, the Toronto Pet User Group, I
and many other readers immediately forecast a world of applications.
Mandala, from The Vivid Group of Toronto, offered a completely GUI-based
authoring environment in which to create applications controlled via the
video input. I.e., you saw yourself - or at least your silouhette (sp?) -
on screen, and it was the control object. You could program any kind of
interaction based on contact, direction, velocity, etc. for real time
interactions with a virtual environment, and you could address and send or
receive from any device on the system - serial, parallel ports, etc., as
well as the Amiga's logical devices - which could be just about anything,
all from within the Mandala GUI. The maximum interactive frame rate was
about 14, but that was good enough for a lot of applications.
One of the applications that a local user - a Mr. Salcedo, as I recall -
developed here in Orange County, CA was to use the drum set that came as a
demo app with a quadrapleagic kid at the Blind Children's Center. This kid
could move one finger on one hand. Using a camera with a macro lens, he
could now play the drums.
During this period, I owned an amazon parrot, which ultimately died from
neglect after pulling out all its feathers and becoming quite neurotic and
hostile. This was my fault, but I just couldn't be there. One of the
obvious applications which I thought of for Mandala was pet training and pet
games for animals like parrots who need constant interaction. It also
occurred to me that a parrot network could be set up by which parrots could
call each other up for company, although I don't think that there was any
off-the-shelf software available then to support such a project. (I did go
so far as to digitally record my parrot's flocking cry and would sometimes
set one of my Amigas to play it back at random intervals when I was not
there.)
Unfortunately, Mandala and the necessary hardware to run it were expensive
and I couldn't afford it. (Vivid originally offered the software quite
cheaply, but then kept raising the price, ultimately to $1,000, and then
discontinued it as a programmable system and went to selling $40,000
stand-alone compiled apps, meanwhile working on developing patents for
individual techniques. Later Intel apparently entered into some kind of
arrangement with them as well.)
By the time I did get CyberScape, a Mandala clone, my parrot had departed
for the great flock in the sky. I did write some articles about various
applications for Videoplaces, which appeared in various Amiga magazines and
later on a Web news service. And I discussed my parrot or pet training
applications with numerous individuals, including pet store owners and vets
over the years.
Since then, CyberScape has appeared for the PC platform running a full 30
frames per second interaction - requiring a MATROX Meteor ($500) card,
unfortunately, and Reality Fusion, the first company formed under Guy
Kawasaki's Garage.com, has a product that overcome's one of the major
problems with previous products - backgrounds.
Other personal computer Videoplaces required a blank background or other
tricks to extract the control object for edge detection. Reality Fusion's
product, which has appeared as compiled demo apps on various PC camera
packages from Intel and Logitech, uses some sophisticated processing to
extract the moving actor from any random background.
My understanding is that the algorythm is based on wavelet compression,
which is how I thought it should be done, and which I suggested to the
author of CyberScape some years ago. However, it does really drag the
system down to add the additional layer of processing, based on the store
demos I've seen so far. The actual interactive frame rate appears to be
about 1 or 2 per second on a 300Mhz PC. The hardware will catch up,
however, now that the basic problem has been hacked.
The "OC Register" yesterday carried a story about a researcher at MIT's
Media Lab - Irene Pepperberg - who had trained a parrot to surf the web,
using a joystick type interface into which it inserts its head. The article
cited the problems that many parrot owners have with keeping their birds
from pining away from loneliness. This system allows a parrot at home to
call up the owner for a chat, or to visit with other parrots on the web.
The Media Lab website has further information on previous research done by
Ms. Pepperberg with parrots.
The odd thing, as it struck me, is that the Media Lab is allegedly one of
the hotbeds of VideoPlace development. Scientific American's cover story
some years back was about their "smart home" project, involving some 60
researchers who were using - in about 90% of the cases described -
videoplace technology for interactions between the home
computing/entertainment/control/automation/alarm/communication system and
the inhabitants. So why didn't Ms. Pepperberg use any of that technology
with her parrot?
There are other oddities connected with the Media Lab's videoplace
endeavors. Myron Kreuger holds the patent on VideoPlace, and he literally
wrote the book on its use, potential and problems to be overcome. He has
done decades of professional research on the technology, including thousands
of experiments in the psychology of interaction. Yet the Scientific
American article never cited him.
I used to cover this stuff for various magazines, as I mentioned above.
When I would go to SIGGRAPH, Mandala would be there in some incarnation, and
MIT would always have a videoplace from the Media Lab. However, the MIT
stuff was generally a total dog. Despite all the hype, as in the Sci. Am.
article, it appeared that they couldn't even come close to Mandala in actual
performance. According to Myron, they were lucky to get 1 or 2 frames per
second, even though they were typically running very powerful hardware - SGI
or the like. He was also quite bitter about the fact that they never gave
him credit for inventing the technology, and completely ignored his patent.
(His actual stated opinions were quite a bit more harsh than this, but I
don't want to misquote him.)
So, this morning I called the Media Lab from work, with idea of having a
friendly chat with Ms. Pepperberg. I wondered why she wasn't using
videoplace, but I also had a list of other related questions and
suggestions. (The security systems manufacturing company for which I work
is planning on getting into CCTV and the smart home stuff at some point in
the future, but they are just looking around at this point, but this at
least gave me a legitimate excuse to call.)
When I got the Media Lab, I was informed that Ms. Pepperberg was travelling,
but I could leave a message. However, when I mentioned that I was calling
from work, I was informed that in that case I should not expect a call back,
as Ms. Pepperberg was contractually bound not to discuss her research with
anyone except the companies providing the funding for the Media Lab. Of
course, a lot of that funding, I'm sure comes from your and my tax dollars,
as well.
Bottom line: As I have suggested for about a decade now, the VideoPlace
stuff as well as a host of other applications - many of them quite obvious
to anyone generally knowledgeable in the field - are being quietly
copyrighted or patented by the big players.
Five years or so from now, when the home level system finally gets past the
PC architecture / Windose bottleneck and Myron's patent expires, we'll see a
huge influx of proprietary apps from the big players and the lone inventor -
the Wozniak or Bushnell - or the small startup company will find themselves
completely shut out. Even people like Myron, who has the very broad patent
on the entire field, will probably find, as he has, that the costs of
enforcing a patent can be prohibitive for an individual or small company.
So, I'm looking for a good place to locate my Free Idea Factory site
meanwhile, where I will solicit people to post great ideas, for which they
would get credit for all posterity, but by which they would be rendered
permanently unpatentable. If I do it right, it will be impossible to
ignore, and all the would-be patenters will have to check there first.
(Maybe I should patent this idea first?)
Suggestions?????
________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 15:30:06 MST