10/27/97
During the classes held this week we discussed the different kinds of input technolgies that Virtual Reality uses. Before we began the lecture on input technologies we concluded our talk about .nff files. These files are the files that are loaded into our Virtual Environments to give us the images that we see. The main point that I took from our talk on .nff files this day was that the files are designed very much like a VRML file with a list of points that make up a polygon. These points are then called on to be created, and the order that the points are called matters.
THE GLOVE
The first
input technlogy we discussed was the glove. The glove is a, well
glove, that measures the amount of bend in the users hand. This bend
could be used to cause a movement on the screen. For example bending
all the fingers usually means that the user has grabbed an object on the
screen. Bending one finger and changing the direction can cause movement
on the screen in whatever direction the finger points. These functions
are not set reactions to the users movement, in other words programmers
can alter the scheme and give a whole different list of commands for input
movements.
EYE TRACKING
A
lot of technologies that Virtual Reality uses were first conceived for
use by people with disabilities. Eye Tracking is an example of such
a technology. There are two technologies of Eye Tracking that are
used. The first technology uses the movements of the eye to guide
the computer. The basic theory of this first technology is an infared
beam is reflected off the eyeball and the computer reads the corneal reflection
to know where the eye is looking. A good example could be when the
desktop is on the computer and you want to open up the Recycle Bin.
With the reflection technology you would focus your eyes at the Bin.
It would recognize this command and open up the bin. You could perform
whatever functions you like from this point on using the same movements
towards the menu at the top of the open Recycle Bin. The second kind
of eye tracking provides less freedom of movement. The computer has
electrodes attached to the temples of the head and images that the eye
sees are sent to the computer through impulses. So, to open the Recycle
Bin the same movements would be needed, but this time it is reading the
stare through the impulses the eye is sending instead of bouncing infared
beams around. The benefit to this technology is that people with
contacts can use it. The problem with all eye tracking technologies,
as of right now, has been called the Midas Touch problem. Basically
Midas Touch means that the eye constantly moves around and if what you
look at is supposed to do something than a lot of things you don't want
to do anything will be reacting to normal eye movements. Glancing
for example is a hard thing for eye trackers to compensate for.
GESTURE RECOGNITION
Basically
this input technology reacts to the users gestures. Different gestures
are interpreted to have the computer do different things. One example
is a thing called Vactors. Vactors is a technology that reads the
users facial expression and, I believe outputs the expression.
SPEECH
There
are two approaches to training a computer to recognize speech. The
first is simply training the computer to recognize the users speech patterns.
The second is using multiple persons speech and when the user tries the
program the computer takes the sound and matches it against a normalized
template. The benefit of using speech is it frees up the use of the
hands completely. There are some problems with this idea though.
If the computer is to be trained to the user, it takes a long time to configure
the computer to recognize every word. There is also the problem of
double meanings, or words that sound alike, but are spelled differently.
Background noise, voice inflections, and homonyms are all problems.
10/29/97
VISUAL PERCEPTION
We discussed a lot about how we as humans perceive objects. A lot of the things brought up were
This page was created: 11.18.97