Reproduce “Adapting Stereo Vision From Objects To 3D Lunar Surface Reconstruction with the StereoLunar Dataset” 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
Clementine Grethen, Simone Gasparini, Geraldine Morin, Jeremy Lebreton, Lucas Marti, Manuel Sanchez-Gestido
The authors introduce LunarStereo, a synthetic lunar stereo dataset, and fine-tune pretrained MASt3R for terrain reconstruction. It contains over 50,000 pairs rendered from South Pole LOLA elevation data with modeled reflectance and sunlight. On synthetic nadir tests, slope-map Pearson correlation rises from 0.21 to 0.80. This measures slope-pattern agreement; testing uses the training DEM, while Chang’e-3 results are qualitative.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.