Computer vision master’s program could address some industry ills
The University of Central Florida claims it is preparing the first master’s program in computer vision to be offered by a public university in the United States. Demand for face biometrics skills alone should lure students into classes.
Already the home of the Center for Research in Computer Vision, the school expects to open the 30 credit-hour degree this fall. Students will do hands-on research with center faculty.
The university boasts that its center had advanced the way mass-crowds are counted, and has created pattern-recognition software capable of identifying lung cancer with 95 percent accuracy. Researchers also have written algorithms that reportedly can differentiate between COVID and the flu in lungs.
An article in Verdict, a magazine published by UK technology analysis and consulting firm GlobalData, teases apart the technology and implications of computer vision. The editors point out what should be obvious but too often is not.
Computer vision is comprised of four categories: machine vision, object recognition, video recognition and, naturally, image recognition. That is an important note for would-be investors and those interested in buying systems.
A self-inflicted wound like biased programming in facial recognition does not necessarily drag down the dynamics of, say, object recognition.
And, in fact, gender and racial bias would be less of a hindrance to acceptance if the biometrics industry could manage its own and market expectations, according to the article.
That might be a course worth exploring.