A NEW IMAGE-TO-VIDEO MODEL

Summary from the TLDR AI newsletter

Researchers have developed image-to-video diffusion models that generate realistic motion transformations from static images, addressing limitations in traditional methods like affine transformations.