Reproduce “Loop Closure from Two Views: Revisiting PGO for Scalable Trajectory Estimation through Monocular Priors” 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
Tian Yi Lim, Boyang Sun, Marc Pollefeys, Hermann Blum
2GO refines robot trajectories without reconstructing a landmark map. It uses image similarity and checks to connect revisited keyframes with scale-free or metric pose constraints, then optimizes them with odometry in a sparse pose graph. On Spot data, it reduced RMSE ATE from 2.44 m to 0.66 m; it still requires images and odometry and does not provide 3D geometry.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.