Where Do We Go From Here?
So Much To Do, So Little Time:
- Read street signs to people walking around.
- Read newspapers and magazines aloud to people.
- Read colors ... tell a blind person:
The sky is slate grey today with big
black rain clouds on the horizon.
(That's the weather all the time here in Seattle)
- Recognize people's faces ... tell a blind person
"You're looking at Matt right now" or
"You're looking at a green Audi Quattro right now" or
"Your bus, the #54, just pulled up to the bus stop.
- Use "super-resolution" techniques to take an image
that's far away (too far to see), and make a
bigger, clearer copy on a screen close to the
user, so the user can
can see the details of the picture, even
if it's hundreds of feet away.
- Make all of this work, and more, on a digital
video feed, which means the user can be
walking around downtown with a digital
camcorder and have all of the machine
vision techniques work on the video stream
they're currently shooting.