www . AmericanSecurityToday . com 2022 CHAMPIONS - Edition 66
Machine Learning has rede- fined Artificial Intelligence ( AI ) in general and Computer Vision in particular .
Computer Vision used to be ar- tisanal , meaning it was more art than science . You wanted a computer to look at a camera and detect cars ? You had to ex- plain to a computer what a car looked like first … and that ’ s harder than you might think .
Now , with Machine Learning , you tell the computer , “ Okay , here are 10,000 pictures of cars and 50,000 pictures of not cars ( boats , people , buildings , mailboxes , etc .),” and the ma- chine learns on its own .
Machine learning has been the equivalent of putting Comput- er Vision on Russian steroids .
Computer Vision is now so pow- erful it can do things it couldn ’ t do before . You can have a com- puter watching all your cam- eras , all the time , cost-effec- tively .
It can locate people where they shouldn ’ t be and / or car- rying things they shouldn ’ t have - at their precise loca- tion in real-time .
With all these improvements , though , Computer Vision still can ’ t replace human intel- ligence . Computers lack the ability to detect nuances in behavior . Computer Vision still generally lacks what , in Machine Learning circles , is usual- ly called “ Persistence .”
9