Intel makes GTA V Photorealistic using Machine Learning

[WYP]

Could AI enhancement deliver the future of gaming realism?



Read more about Intel using Machine Learning to make GTA V more photorealistic.

 

[Dicehunter]

I know it won't happen but it would be great if Intel released a full AI modded texture/lighting pack for GTAV.

 

[KingNosser]

Pretty impresive for Intel, but unless it comes to the public it's imho a wasted effort other than oh we have a nice tech demo, sure they will have learned something but no real point learning if you don't put it to use.

 

[ImprovizoR]

It's a pretty impressive proof of concept, and the potential implications are astounding.


Think about this, for instance. Rockstar goes around taking a lot of photos and videos of cities in order to create the fictional version. Theoretically, they could partner-up with Intel and use this method during the development of GTA VI in order to make it more photorealistic. Same goes for other devs.

 

[Dicehunter]

It's a pretty impressive proof of concept, and the potential implications are astounding.


Think about this, for instance. Rockstar goes around taking a lot of photos and videos of cities in order to create the fictional version. Theoretically, they could partner-up with Intel and use this method during the development of GTA VI in order to make it more photorealistic. Same goes for other devs.

This likely won't be used for anything that we can get our hands on, Same with Nvidia, They've showcased some really cool graphical tech over the years that simulates fur, Fire, Water etc... and it could be used in games, Is it ? Nope it's unused and forgotten.

 

[grec]

This likely won't be used for anything that we can get our hands on, Same with Nvidia, They've showcased some really cool graphical tech over the years that simulates fur, Fire, Water etc... and it could be used in games, Is it ? Nope it's unused and forgotten.

To be fair, really detailed environmental or character demo tech often looks nice and is easy to do for pre-recorded footage, but is usually a pain to get working in realtime while often not being worth the performance penalties in motion

With this, the performance penalty on AI accelerated hardware should be pretty minimal, and fairly flat across the board, while integration could be somewhat tacked on after the fact.

The potentially to use this to create multiple "themes" for a game- Dark, colourful, realistic, ect, or say historical-footage themed filters, is very interesting imo, particularly for say racing or simulator games where artistic styling isn't an intrinsic aspect of the game.

And with something like this, maybe we don't really even need official support, it seems pretty mod-friendly

 

[NeverBackDown]

To be fair graphic technology takes years before it gets into consumer hands. They are generally pushing the envelope of what's capable, it's not feasible to give it to the masses immediately as it still needs to get integrated into software on a production scale and then hardware that supports the new feature sets and horsepower to run it real time at ideal performance.

It's basically like how RT took quite a while to get into any games. It was basically buy RTX for DLSS 1.0 for 2 years.

 

[demonking]

I thought this was quite impressive on a game like gta 5. Its quite old now and is limited by its age and its a procedurally generated sandbox so this needs quite a bit of processing power to do.
Picking something that had a much smaller sandbox environment that is loaded pre entry (like a racing game where the whole sandbox is loaded) would have been a much simpler approach so well done intel.
I'm not sure if the current hardware we have these days is enough to support this but we're not too far off, IMO the biggest hurdle would be storage. With most AAA games now taking well over 100gb with just their standard textures and models, adding in enough data for the AI to work from effectively is going to push the storage requirements up astronomically.

 

[grec]

AI training and inference are quite different tasks, the inference for this doesn't rely on anything in-engine besides some intermediate output buffers, so game assets, complexity, ect shouldn't have much impact on ease or size of deployment at runtime (Though the structure of a particular game engine probably would), just the training that would have to be done to get the inference model up to scratch before deployment

 

[AngryGoldfish]

Was anyone else reminded of this video when watching that Intel demonstration?

https://www.youtube.com/watch?v=RXJKdh1KZ0w

 

[Warchild]

Was anyone else reminded of this video when watching that Intel demonstration?

https://www.youtube.com/watch?v=RXJKdh1KZ0w

All made sense to me. however...
I think they forgot to include the quasi-flange-bananabot-inductor-thunderpant that powers the national grids 35 chromosomes consisting of strawberry blancmange.