One scene, two matched prompts: the image prompt makes the frame, the video prompt animates it — character and aspect identical in both.
The core feature of the Dreamer pipeline: a matched image + video prompt from a plain language scene in one step. Step 1 generates the image, Step 2 animates it. For dialogue instead of text-to-video, lip-sync to the held still.
More in the lexikon
Related terms
Test your knowledge
Quiz
1. Zu welchem Department gehört „Image-to-Video Matched Pair"?