Question

A marketing team wants to produce a short promotional video clip that animates a still product photograph they supply, using a text prompt to describe camera motion and lighting changes. They are building this on Azure AI Foundry and need to select the correct generation approach so that the supplied product image is used as the visual starting point rather than generating scenes from scratch. Which approach should they use?

Accepted Answer

Use a video generation model with an image-to-video (reference media) input, passing the product photo as the seed frame along with the text prompt describing motion and lighting

Answer

Use a text-to-image model to regenerate the product photo for each frame, then stitch the frames into a video file manually

Answer

Use Azure Content Understanding to analyze the product photo and return captions describing the desired motion

Answer

Use an image-editing inpainting workflow with a mask to add motion blur to the product photo