Meta AI: AI-generated video is coming

shakiroslann -

Meta AI is an AI lab that belongs to Meta Platforms Inc has announced a new AI system called Make-A-Video that lets people generate text prompts into brief video clips.

Make-A-Video gallery screenshot.
A brief video generated by Make A Video with "A confused grizzly bear in calculus class" prompt.

Make-A-Video is built on Meta AI's recent progress in generative technology research. The system learns what the world looks like from paired text-image data and learns how the world moves from unsupervised video footage.

Another thing the system is able to do is to conceptualize a video from a single image or pair of images, even more, you can add extra creativity to your video by creating variations of your video based on the original video. You can see more in their gallery.

An example of a classic piece of art from Make A Video that is going to be generated.
A brief video that has been generated from a single image.

It's a bit similar to other text-to-video models but this one they're doing it in a slightly different way with unlabelled video data.