[midlightdenight]

It looks like DALLE is just performing inpainting on pieces of the image outside the original frame.

The zoom effect is scaling the original frame larger while keeping the frame, then performing image to image generation on it, or using the newly defined image in the frame and sending it back through diffusion (at least that’s my guess).

It’s a different process than what DALLE has since inpainting does not overwrite already generated pieces. Stable diffusion can also do inpainting.

You can make sliding images this way. Slowly translating the image out of the frame and then filling in the blank space with inpainting.