Reproduce “One2Scene: Geometric Consistent Explorable 3D Scene Generation from a Single Image” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Pengfei Wang, Liyi Chen, Zhiyuan Ma, Yanjun Guo, Guowen Zhang, Lei Zhang
One2Scene turns a single image into an explorable scene by generating a panorama, building a shared 3D Gaussian scaffold from six views, and using it to guide new views. On a 40-scene evaluation, it reports CLIP-I 89.95 versus 88.70, and CamMC 0.389 versus 0.460 for SEVA+Anchor; results support improved consistency in tested settings, not perfect geometry.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.