This is a video discussing and demonstrating a method for rendering a much more photorealistic version of Grand Theft Auto V using an artificial intelligence powered convolution network that produces images frame-by-frame at speeds fast enough to be run at interactive (read: playable) rates. Some more info that’s so far over my head it might as well be the International Space Station:
We present an approach to enhancing the realism of synthetic images. 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 artifacts 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. We also introduce multiple architectural improvements in the deep network modules used for photorealism enhancement. We confirm the benefits of our contributions in controlled experiments and report substantial gains in stability and realism in comparison to recent image-to-image translation methods and a variety of other baselines.
No clue what any of that meant, but the results are almost unbelievable. I mean coupled with that technology that messes with your balance to make you feel like you’re actually in a vehicle, soon you won’t be able to tell if you actually ran over that person or if was just a very realistic video game. “There’s blood on your bumper.” Well I guess that answers that.