Reproduce “PCPNET: Learning Local Shape Properties from Raw Point Clouds” 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

Paul Guerrero, Yanir Kleiman, Maks Ovsjanikov, Niloy J. Mitra

PCPNet estimates local geometric properties of raw point clouds, including normals and curvature. It processes neighborhoods around target points, using learned features and optionally multiple scales to reduce reliance on hand-tuned local fitting. The authors report lower unoriented normal errors than compared methods across tested noise levels, but performance is weaker under non-uniform sampling because training used uniform samples.

Summary by alphaXiv.

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