XO avatars

[Edward Cherlin]

On 3/30/07, Dan Williams <dcbw at redhat.com> wrote:
> On Fri, 2007-03-30 at 12:42 -0700, Hal Murray wrote:
> > > Is there any thought to composing the avatars from a set of components?
> >
> > Why not use the camera?
> >
> > I've worked with a "faces" mail system that tossed up a new icon whenever you
> > received new mail from somebody.  We had a picture for everybody in the
> > group.  (and a big collection of corporate/university logos for when the
> > individual wasn't in the database)
>
> We're going to have a "buddy picture" (which in XMPP is an "avatar") for
> each person that will be taken by the camera on first boot.  It will be
> changeable.  It'll be available from roll-overs on each XO icon wherever
> you see that icon in Sugar.
>
> Dan
>
> > I forget how big they were.  Google found one mention of a utility to
> > compress/decompress 48x48 face pictures.
> >
> > I remember being impressed at how well it worked with such a small space.
> > Mostly it was instant recognition with just a quick glance.  There were a few
> > people in the group with pictures that were similar enough to cause me
> > troubles.  For those I had to take a long look.

Yes, face recognition was heavily optimized by evolution long before
the emergence of humans. They say that you can tell you're getting old
when everybody you meet looks like someone you already know, which
fits nicely with the evolutionary model.

There is a study that says that all faces can be composed from 77
eigenfaces, which would allow very good compression.


-- 
Edward Cherlin
Earth Treasury: End Poverty at a Profit
http://wiki.laptop.org/go/Earth_Treasury
WIRE AFRICA  http//www.wireafrica.org/
http://www.linkedin.com/in/cherlin