Read the arXiv paper to understand the 3D reconstruction approach.
Explore the project's interactive demo website to see example outputs.
Cite this work using the citation information provided in the README.
Check back later once the code is released to try reconstructing your own photos.
| shirleymaxx/rest3d | abgcto/hey-claude | ernie-research/nava | |
|---|---|---|---|
| Stars | 62 | 62 | 62 |
| Language | — | Swift | Python |
| Setup difficulty | easy | easy | hard |
| Complexity | 1/5 | 2/5 | 5/5 |
| Audience | researcher | developer | researcher |
Figures from each repo's GitHub metadata at analysis time.
No usable code is available yet, only the paper, citation, and a web demo are provided.
REST3D is an academic research project from Carnegie Mellon University that reconstructs an interactive 3D scene from a single ordinary photograph. The goal is to take one casual image, such as a photo of a room or outdoor environment, and produce a 3D representation of that scene that is both visually consistent with the original image and physically stable, meaning objects sit properly on surfaces rather than floating or clipping through each other. The project was published as an academic paper in May 2026. The repository currently holds the citation information and links to the paper on arXiv and to an interactive demo on the project website where you can explore example outputs. The actual code has not been released yet. The README notes that code is coming soon. There is not much more detail available in the repository at this stage. The description says the system starts from a single image and produces an interactive 3D scene, but the technical approach and any usage instructions are in the paper and project page rather than in this repository.
An academic research project that reconstructs a physically stable, interactive 3D scene from a single ordinary photo. Code is not yet released.
The README does not state a specific license for this project.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly researcher.
This repo across BitVibe Labs
Verify against the repo before relying on details.