SORA Technical Explanation

Video generation models as world simulators

Explore large-scale training of generative models on video data at OpenAi. Specifically, OpenAi trains text-conditional diffusion models jointly on videos and images of variable durations, resolutions and aspect ratios. They leverage a transformer architecture that operates on spacetime patches of video and image latent codes. Their largest model, Sora, is capable of generating a minute of high fidelity video. Our results suggest that scaling video generation models is a promising path towards building general purpose simulators of the physical world.

Last updated