Smiley Toolkit

[ism]

The Smiley Toolkit (http://snookie.apostasy.org/~agrant/semiotics.html) is an interesting method of attempting to convey expressions through machine-mediated communication. A person's mood is quantified into 6 categories, transmitted as raw data, then combined on the other end to create an expression matching the user's mood.

[kimpossible]

Whoa... if you leave everything else alone and turn up Anger and Disgust full blast you get neutral/mellow look.

[mr. x]

whoa man that's pretty cool, a lil flawed here and there but imagine, a few years from now you could do that with the Gman or something (for you non-Halflife fans the Gman is the briefcase carrying suit who continually thwarts our hero at every turn)

[ism]

I'm thinking of the reverse situation. While IDing a person based on matching pictures to a database is currently flaw-prone, image processing is good enough to get a general idea of a person's facial expression. Semiotics, dealing with any type of sign, could apply to a person's face, as the expression is the signifier. If expressions could be quantified, there could be some interesting applications.

The most obvious one is figuring out a person's mood. This could be broken down into more specific acts such as lie detecting. Baseline images would be taken and variations could then be compared. In combination with other methods, could be more accurate (personally I'm not keen on the polygraph and not sure how much more successful this would be)..

Then you could cross over into medical semiotics and try to figure out what kind of pain or sensations a patient is feeling. By taring the system by gauging relative reactions to a set of stimuli, pain that cannot be normally quantified can be. For example, pain can be categorized as dull, throbbing, sharp, and in intensity. With such a system, these factors can be quantified and matched against a database of sensation, and perhaps a diagnosis can be made, or possibilities narrowed down.