Meet Stable Virtual Camera: a multi-view diffusion model that turns 2D images into 3D videos. You can use only one image or up to 32 images. It supports various trajectories, including orbit, dolly zoom, and spiral. I quickly tried a Midjourney image with it, and the results were interesting (see above video).
As explained on Stability.AI, this model creates 3D videos with realistic depth and perspective. It supports square (1:1), portrait (9:16), and landscape (16:9) videos. It ensures consistency in up to 1000 frames.
[HT]