EVNow
Well-Known Member
This is the "end to end" training. Tesla isn't doing it - yet.This same thing can be used to help with "what is the other driver going to do?" Show it many instances of that situation, and look at the outcome of what the other driver did. And repeat this millions of times. You end up with a pretty good idea of how the average human is going to react in a given driving situation.
They use "planner" a procedural code written to figure out the action. They use NN only for perception at this time. The planner figures out the route by running an optimizer that picks the most optimal route based on several parameters (and weights).