PAC-MAN Reverse Engineered with NVIDIA GameGAN AI

PAC-MAN Turns 40, Gets AI Makeover

In December of 2018 we saw the first proof-of-concept for an AI-powered virtual world that had gaming implications – some day. Well, today NVIDIA is showcasing a new AI-powered gaming concept, and it might just be familiar to you.

Using an AI model called NVIDIA GameGAN (a “GAN” being a generative adversarial network) created by NVIDIA Research, some 50,000 episodes of PAC-MAN were used “to produce a fully functional version of the classic without an underlying game engine”.

The result? (Video embedded below)

GameGAN is the first neural network model that mimics a computer game engine by harnessing generative adversarial networks, or GANs. Made up of two competing neural networks, a generator and a discriminator, GAN-based models learn to create new content that’s convincing enough to pass for the original. You may have experienced GANs by using NVIDIA GauGAN, a deep learning model developed by NVIDIA Research that turns rough doodles into photorealistic masterpieces with breathtaking ease.

GameGAN could accelerate the creative process for game developers, as it could be used to automatically generate layouts for new game levels. GameGan can also be used by AI researchers to more easily develop simulator systems for training autonomous machines, where the AI can learn the rules of an environment before interacting with objects in the real world.

So, when can we play this? NVIDIA says that they “will be making our AI tribute to the game available later this year on AI Playground, where anyone can experience our research demos firsthand”. You can read more about this with the research paper from GameGAN available here, and check out NVIDIA’s blog post this morning.