Reproduce “Joint 2D-3D-Semantic Data for Indoor Scene Understanding” 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

Iro Armeni, Sasha Sax, Amir R. Zamir, Silvio Savarese

2D-3D-Semantics provides registered image and 3D data for six indoor areas. It contains 70,496 rendered RGB views, 1,413 panoramas, geometry, labels, meshes, and point clouds. The authors annotate point clouds, then transfer labels to meshes and images. A point-cloud detection baseline reports mean AP 49.93, versus 41.87 without global geometry; this tests detector features, not added modalities.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.