Identify vehicles, pedestrians, traffic signs and objects in video footage to train autonomous and semi-autonomous driving systems.