←back to thread

Full Autopilot in GTA Using TensorFlow

(littlemountainman.github.io)
486 points littlemtman | 2 comments | 12 May 20 15:17 UTC | HN request time: 0.538s | source
Show context
braunshedd ◴[12 May 20 16:12 UTC] No.23155670[source]▶
I'd love to see this done in Forza, although I'm sure that's optimizing for the wrong kind of driving behavior (lol)

Amazing work!

replies(3): >>23155878 #>>23156072 #>>23156658 #
1. mhh__ ◴[12 May 20 16:41 UTC] No.23156072[source]▶
A'I' in racing games is already a cracked problem I think (unless you mean forza horizon?).

The AIs in Rfactor2 are even pretty good wheel to wheel on some tracks - definitely a lot better than online rookies who are quick but blind. The only noticeable thing is that they don't learn over the course of a race. They seem to be programmed to make mistakes realistically, i.e. if the car loses grip in the wet they don't lock back onto the track.

replies(1): >>23156192 #
2. bowmessage ◴[12 May 20 16:51 UTC] No.23156192[source]▶
Sure, but this is a different kind of agent - one that takes screen frames as input, rather than one that has direct access to game state.