It’s pretty much a given that Google Glass will come out with some sort facial recognition software. Now a team of Duke University and University of South Carolina researchers is working on adding another piece of software that will allow your face-attached computer to recognize you in a crowd based on what you are wearing.
The idea behind InSight is that the app will be able to pick out your friends based on their fashion sense in a crowded place (such as a mall) or from a particularly far-off distance that makes facial recognition unfeasible.
To identify anyone, InSight first needs to record the person’s “fashion fingerprint” based on the outfit they are wearing. It creates this “fashion fingerprint” by snapping pictures of them and analyzing the patterns and spatial distribution of colors on the subject (your friend) to create a file called a “spatiogram.”
So far, the app has been tested by 15 volunteers and it managed to correctly identify people 93 percent of the time. Also, if you’re worried about your privacy, the fashion fingerprint only works if you are wearing the same outfit that you wore when the system initially identified you. If you don’t want to be spotted by your clothes, all you have to do is change them…just like the Pretender. Whew.
Researchers unveiled the InSight system at the HotMobile technology conference held in Georgia last week. Duke University students are still currently working on perfecting the technology, with the help of funding from Google.
If you want to read more about the InSight system, the Duke research team has put up a paper on its work.
This story, "Google Glass could identify you by your clothes" was originally published by TechHive.