For humans, identifying items in a scene — whether that’s an avocado or an Aventador, a pile of mashed potatoes or an alien mothership — is as simple as looking at them. But for artificial ...