ALEXANDRIA, Va., Sept. 15 -- United States Patent no. 12,737,937, issued on Sept. 15, was assigned to GDM Holding LLC (Mountain View, Calif.).

"Video generation and encoding using machine-learning models" was invented by Sander Etienne Lea Dieleman (London), Junyoung Chung (London), Marc van Zee (Humlebaek, Denmark), Shlomo Fruchter (Ness Ziona, Israel), Jeffrey Donahue (London), Yutian Chen (Cambridge, Great Britain) and Ankush Gupta (London).

According to the abstract* released by the U.S. Patent & Trademark Office: "This specification discloses systems, method and apparatus for performing operations relating to video generation and/or video encoding using machine-learning models, such as latent diffusion models. In some aspects, latent...