Reproduce “Single View Metrology in the Wild” 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
Rui Zhu, Xingyi Yang, Yannick Hold-Geoffroy, Federico Perazzi, Jonathan Eisenmann, Kalyan Sunkavalli, Manmohan Chandraker
ScaleNet estimates camera orientation, field of view, camera height, and object heights from a single image, using people and cars as scale cues. It predicts object boxes and heights, then uses perspective reprojection and refinement to align them with the image. On KITTI-mulCat, its mean absolute camera-height error is 0.1222 m, versus 0.1519 m for PGM, under a flat-ground assumption.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.