Reproduce “AlphaTablets: A Generic Plane Representation for 3D Planar Reconstruction from Monocular Videos” 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
Yuze He, Wang Zhao, Shaohui Liu, Yubin Hu, Yushi Bai, Yu-Hui Wen, Yong-Jin Liu
AlphaTablets reconstructs 3D planes from posed monocular video using rectangles with color textures and transparency maps, allowing irregular boundaries on continuous surfaces. The authors initialize tablets from image regions, depth, and surface directions, optimize them against video, and merge neighbors. On ScanNet, it reaches an F-score of 0.456 versus PlanarRecon’s 0.372, though its Acc error is higher.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.