Can We Make A Better FRT Facial Recognition Technology Software Model?

Today’s facial recognition systems are getting better. I suppose soon you’ll be able to upload a picture of your cat or dog that’s lost and the search engine will search local animal shelter, pound, veterinarian offices, and all of the public digital video feeds from street corners, UAVs, satellites, police cruisers, buses, postal jeeps, etc., and if any system has found it or anyone has uploaded a picture of their pet to Craig’s List, Facebook, or some other social network marked “found” then you’ll be alerted in 0.0882342 seconds. See that point?

Have you even told someone that you aren’t that good with names but you never forget a face? Have you ever heard someone else say that? I have. Okay so, let’s talk shall we?

In Medical Express Online News on November 26, 2012 there was an interesting article titled; “Study: To get the best look at a person’s face, look just below the eyes,” which stated;

“They say that the eyes are the windows to the soul. However, to get a real idea of what a person is up to, according to UC Santa Barbara researchers Miguel Eckstein and Matt Peterson, the best place to check is right below the eyes. Their findings are published in the Proceedings of the National Academy of Sciences.”

Maybe there is a reason why humans do this and concentrate on that part of the face? Maybe, we could use this reality that social evolution has provided us; that our scientific researchers have discovered? Maybe we need more areas of the FTR software mathematical grid in those areas for greater definition and scrutiny? Maybe we can train our FRT software to do a much better job with all of this you see? It doesn’t take very much work to put an “X” inside each of those grid areas, giving us another point to measure inside, or a grid within those grids in that region of the face.

If we doubled up in those regions or even tripled up our grid, we’d have much higher definition in the areas that count. Now, I know what you are thinking, the human mind and organic optical system is much different than the FTR system, but, I wonder, further a human will be looking at the results once the FTR shows the matches to verify them in many cases, thus, shouldn’t it get those particular points exact in that case?

Now then, if you’d like to discuss any of this at a much higher level or play around with the mathematics involved or the programming of such a next generation FTR the let’s talk. Until then, please consider all this and think on it.