From an AI insider, In computer Vision AI this is called “detection”. It can recognize objects in a scene and it attempts to segment them, hence the bounding boxes around recognized objects. This isn’t the most advanced for of computer vision using deep neural networks... But it is always pleasing to show to a crowd.…

Yeah I’m sure people will hate on this, but what’s revolutionary about this compared to driver assistance systems from Chrysler and Mercedes Benz is the machine learning. If it can do this well with such little data - imagine after a year of machine learning with millions of more miles and use cases.