The term “face recognition,” or “facial recognition” is pretty much self-explanatory in as much that it is exactly what it sounds like – technology that is designed to recognize faces. The biometric technology has many uses, from helping law enforcement to “catch their man” to helping your business to target its audience. Face recognition from image and face recognition from video works by mapping certain points of a person’s face and making it possible for that face to be verified when it is seen again, much the same way that we can map a location so that it can be recognized.
Face Recognition’s Rise to Fame
According to a report from Science ABC, face recognition rose to fame in the United States as a way for the U.S. Department of Defense to recognize terrorists and criminals when they attempted to cross the border into the country. Funds and grants for research were granted to some of the country’s top scientific experts in the biometric field in exchange for their aid in getting this project off the ground. The ability to identify and apprehend criminals coming into the United States through airports, train stations, border crossings, and other means was quite valuable to the DoD.
Face Recognition History – The Early Years
Of course, face recognition was developed much earlier than the 90s. As Wikipedia explains, the concept actually got its start from face recognition technology pioneers Woody Bledsoe, Charles Bisson, and Helen Chan Wolf. These three began their work of using a computer to recognize people’s faces during 1964 and 1965. There is not a lot of information about this early work in face recognition on Google because the funding for their project was provided by an intelligence agency that wanted to remain unnamed and would not allow much publicity. Although the scientists were proud of their work, not much of it was published.
Early Issues with the Face Recognition Project
Of course, just as with any project, the first attempts at face recognition came with its own issues. As Bledsoe explained it, the initial approach to making the computer recognize faces was to map out various landmarks on a person’s face, such as the center of the eyes, the corners of the mouth, and such. The computer was to use these maps to mathematically rotate the head to compensate for variations in poses.
As you can imagine, this caused issues when face recognition was put to the test on real subjects. In 1966, Bledsoe noted, “This recognition problem is made difficult by the great variability in head rotation and tilt, lighting intensity and angle, facial expression, aging, etc. Some other attempts at face recognition by machine have allowed for little or no variability in these quantities. Yet the method of correlation (or pattern matching) of unprocessed optical data, which is often used by some researchers, is certain to fail in cases where the variability is great. In particular, the correlation is very low between two pictures of the same person with two different head rotations.”
In our next article, we will discuss more about face recognition, how it works, and face recognition history in part 2 of our article, “What is Face Recognition?”