โ All systems
Diffusion Video Generation
Text-to-video: a denoiser with a time axis bolted on
Architecture introduced 07 Apr 2022
Video generators work like image generators with a sense of time added: the same 'clean up the noise' trick makes a frame, and a motion module keeps the frames consistent so they move smoothly. Feed it a reference face or a pose and it can make a specific person appear to move and act.
InstructionsDataActionsControl / decisionFeedback / logs
๐ Click any component in the diagram to inspect its risks & defensesFollow a request ยท step 1 of 4
โ / โ keys
You describe a clip โ and optionally hand it a reference photo of a face or a pose to follow.