Reproduce “PlanarCityLoc: Breaking Silhouette Ambiguity in LoD Aerial Localization via 3D Planes” 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
Oussema Dhaouadi, Shao Jie Hu-Chen, Jakob Gollreiter, Samir Jusufi, Nikolas Rieger, Caroline Wendlandt, Christian Wender, Jingfeng Mao and 5 more
PlanarCityLoc refines drone poses by matching building planes in an aerial image to surfaces in a lightweight city map. It first aligns image and map geometry using masks, depth, and normals, then estimates pose from 3D-to-2D plane correspondences. On UAVD4L-LoDv2 in-Traj., joint recall at 2 m–2° rises from 36.0% to 75.4%; refinement can reduce recall, as on Swiss-EPFLv2 in-Place.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.