liuyuan-pal/SyncDreamer
A diffusion model that generates multiview-consistent images from a single input photograph for 3D reconstruction.
Not currently ranked — collecting fresh signals.
star history
SyncDreamer is a generative model that synthesizes multiple consistent views of a scene or object from a single 2D image. It uses a synchronized multi-view diffusion approach to ensure geometric consistency across generated viewpoints, enabling 3D reconstruction from monocular input. The project includes inference code, pretrained models, training code, and evaluation tools.
Frequently asked
- What is liuyuan-pal/SyncDreamer?
- A diffusion model that generates multiview-consistent images from a single input photograph for 3D reconstruction.
- Is SyncDreamer open source?
- Yes — liuyuan-pal/SyncDreamer is open source, released under the MIT license.
- What language is SyncDreamer written in?
- liuyuan-pal/SyncDreamer is primarily written in Python.
- How popular is SyncDreamer?
- liuyuan-pal/SyncDreamer has 1k stars on GitHub.
- Where can I find SyncDreamer?
- liuyuan-pal/SyncDreamer is on GitHub at https://github.com/liuyuan-pal/SyncDreamer.