Wonderland: Navigating 3D Scenes from a Single Image
Camera-guided Video Generations Gallery
(Demos are generated from our camera-guided video diffusion model. All the instances are unseen during model training.)
Precise Camera Control over Different Image Prompts using Same Camera Trajectories
Input Images
Camera Trajectory
Input Images
Camera Trajectory
Extensive Exploration over Scenes using Multiple Camera Trajectories
Input Image
Generated Video
Input Image
Generated Video
