GameNGen

Neural model-driven real-time game engine.

CommonProductImageNeural modelsReal-time interaction
GameNGen is a fully neural model-driven game engine capable of real-time interaction with complex environments while maintaining high quality over extended trajectories. It can interactively simulate the classic game 'DOOM' at over 20 frames per second, with its next-frame prediction achieving a PSNR of 29.4, comparable to lossy JPEG compression. Human evaluators only slightly outperform random chance in distinguishing between game clips and simulated clips. GameNGen is trained through two phases: (1) an RL-agent learns to play the game and records the actions and observations from the training sessions, which become the training data for the generative model; (2) a diffusion model is trained to predict the next frame, conditioned on the past actions and observation sequences. Conditional enhancement allows for stable autoregressive generation over long trajectories.
Visit

GameNGen Visit Over Time

Monthly Visits

49160

Bounce Rate

55.48%

Page per Visit

1.2

Visit Duration

00:00:09

GameNGen Visit Trend

GameNGen Visit Geography

GameNGen Traffic Sources

GameNGen Alternatives