Alert 101: Video Analytics

I took a different route to work this morning which had me pass by the Harvard School of Public Health. Big windows along the street revealed a giant television screen with dozens of smaller frames and the back of a security guard’s head viewing them all. I found it odd they’d put the video monitoring room right on the street like that, but even more I thought how difficult that security guard’s job must be. How can you reliably keep track of that many frames and everything going on in each of them?

According to a video released today by ALERT, a multi-university, Department of Homeland Security Center of Excellence housed at Northeastern, you can’t. Our brains are only able to process information from two screens at a time, apparently. (Now I don’t feel so bad about the deep anxiety I experienced when walking by the tower of televisions in the living room of my college dorm.)

Since these days there’s a lot more visual data to be processed than there are naked eyes to process it, researchers are developing video analytics models to help out. It’s a tricky task though, and still in its infancy. But it’s a pretty neat technology as explained here. Enjoy!