GAN Theft Auto is a neural network's attempt to recreate GTA5

Cal Jeffrey

Posts: 3,125   +863
Staff member
TL;DR: A couple of programmers wanted to see if they could train a generative adversarial network (GAN) to create its own version of Grand Theft Auto 5. The result was a somewhat blurry acid trip but is instantly recognizable as GTA5.

Harrison Kinsley, who goes by Sentdex on YouTube, and his partner trained a fork of Nvidia's GameGAN neural network using a black car on a short section of highway in Grand Theft Auto 5. Nvidia loaned them its DGX Station, which is equipped with four A100 80-gigabyte cards to help with the processing. Kinsley explains what they did and demos the results in the video above.

Initial models were very pixelated, but Kinsley improved this with AI-assisted supersampling. While it's not pretty, it's important to keep in mind that this is not GAN-generated footage. It is a real-time interactive demo. Kinsley is driving an AI-created car in a fully AI-created environment.

Kinsley said that since the DGX Station was on loan, his training time was limited. Although the model does try to compute obstacle clipping in some instances, he would have liked to run more collision samples. Kinsley also wanted to see how much of the GTA5 map the GAN could process. However, that would have required hours more training as he incrementally increased the driving distance, which he did not have time to do.

The video shows the neural network did a pretty decent job recreating some unexpected details. For efficiency's sake, one might think that the GAN would ignore shadows and the sun reflecting off the car. To Kinsley's surprise, it didn't. Shadows, lighting, and reflections move more or less as expected. The GAN also created its own elementary physics system after training it by running into things. An example is how another car skews to the right when tapped on the left rear quarter panel. Head-on collisions are not handled as well.

Kinsley has posted the playable demo dubbed "GAN Theft Auto" to GitHub for those interested in giving it a spin. Again, it's just a tiny portion of the GTA map, and it's not like playing a real game. It's more tech demo than anything else, but it is interesting to watch what the model does in untrained situations. That's when things get a little psychedelic.

Permalink to story.

 

Dimitriid

Posts: 941   +1,734
It does look blurry, but the benefit is that these AI can probably do "crunch" dev time without silly needs like having to shower, eat or eventually stop before (or after) having a mental breakdown and quitting the job just to be able to meet arbitrary deadlines set by millionaire executives.
 

noel24

Posts: 746   +915
How much time of This workstation would one need to recreate San Andreas or Vice City in GTAV engine?
 

Vulcanproject

Posts: 1,386   +2,448
This is how all Hollywood movies are created these days anyway.

Input all the most profitable movies from the past 20 years and then the machine spits out a script and casting suggestions, which invariably is Dwayne Johnson + any vaguely attractive female
 

Irata

Posts: 1,663   +2,793
Maybe that‘s how Cyberpunk‘s console versions were developed…would explain a lot.
 

yRaz

Posts: 3,799   +3,911
It's recreating a game engine. Not developing a game with a plot, quests, and npcs. Still not impressed.
Not only are you missing the point, you don't understand what the AI is doing. There is no game engine, watch the video
 

nnguy2

Posts: 308   +610
Not only are you missing the point, you don't understand what the AI is doing. There is no game engine, watch the video
🙄 Game engine = physics model used to replicate the world. I understand exactly whtits doing and there's no need to do it.
 

yRaz

Posts: 3,799   +3,911
🙄 Game engine = physics model used to replicate the world. I understand exactly whtits doing and there's no need to do it.
It specifically said that it isn't using a physicals model. Amd a game engine isn't just a physics model.....
 

nnguy2

Posts: 308   +610
It specifically said that it isn't using a physicals model. Amd a game engine isn't just a physics model.....

"Although the model does try to compute obstacle clipping in some instances, he would have liked to run more collision samples. "

"For efficiency's sake, one might think that the GAN would ignore shadows and the sun reflecting off the car. To Kinsley's surprise, it didn't. Shadows, lighting, and reflections move more or less as expected. The GAN also created its own elementary physics system after training it by running into things. An example is how another car skews to the right when tapped on the left rear quarter panel. Head-on collisions are not handled as well. "

Sounds a lot like a physics model don't it? Of course game engine isnt JUST than physics but for the most part it's how objects interact.


Here's what Wikipedia has to say about a game engine

" The core functionality typically provided by a game engine may include a rendering engine ("renderer") for 2D or 3D graphics, a physics engine or collision detection (and collision response), sound, scripting, animation, artificial intelligence, networking, streaming, memory management, threading, localization support, scene graph, and video support for cinematics."

And it does sound a lot like its an AI trained replicate of GTA V game engine.
 

terzaerian

Posts: 972   +1,416
It does look blurry, but the benefit is that these AI can probably do "crunch" dev time without silly needs like having to shower, eat or eventually stop before (or after) having a mental breakdown and quitting the job just to be able to meet arbitrary deadlines set by millionaire executives.
What are we going to do when not even "learning to code" will gain us steady employment? 🤔