This repository is the official implementation of Reangle-A-Video.
Reangle-A-Video: A unified framework for synchronized multi-view video generation from a single monocular video, without relying on any multi-view generative prior
Full abstract
We introduce Reangle-A-Video, a unified framework for generating synchronized multi-view videos from a single input video. Unlike mainstream approaches that train multi-view video diffusion models on large-scale 4D datasets, our method reframes the multi-view video generation task as video-to-videos translation, leveraging publicly available image and video diffusion priors. In essence, Reangle-A-Video operates in two stages. (1) Multi-View Motion Learning: An image-to-video diffusion transformer is synchronously fine-tuned in a self-supervised manner to distill view-invariant motion from a set of warped videos. (2) Multi-View Consistent Image-to-Images Translation: The first frame of the input video is warped and inpainted into various camera perspectives under an inference-time cross-view consistency guidance using DUSt3R, generating multi-view consistent starting images. Extensive experiments on static view transport and dynamic camera control show that Reangle-A-Video surpasses existing methods, establishing a new solution for multi-view video generation.
| Input video | Generated video1 | Generated video2 | Generated video3 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
- Release code for Dynamic camera control
- Release code for Static view transport