GTA V Becomes Photorealistic Thanks To Machine Learning Tricks By Intel

This just might be a peek into the future of video games as they get closer to photorealism. Intel has been experimenting with new machine learning technology, and using it on GTA V gameplay just made Los Santos a real city. Forget RTX, this is the real deal.

Well, that’s cool, and scary at the same time! While we are still years away from getting something like this to run in real-time while playing a game, the implications of AI enhancements in games is something to think hard about right now. After all, the casual gamer just got their hands on machines with ray tracing support, something which wasn’t feasible even 5 years ago.

Intel provided a brief explanation of the technique on its website-

“The images are enhanced by a convolutional network that leverages intermediate representations produced by conventional rendering pipelines. The network is trained via a novel adversarial objective, which provides strong supervision at multiple perceptual levels. We analyze scene layout distributions in commonly used datasets and find that they differ in important ways. We hypothesize that this is one of the causes of strong artefacts that can be observed in the results of many prior methods. To address this we propose a new strategy for sampling image patches during training.”

You can read the full paper on Intel’s Photorealism experiment here. Meanwhile, if you’re still interested in GTA V, which was released almost a decade ago, you can wait for its next-gen update which was promised back during the PS5 reveal, due out sometime this year.

Original Article