Reproduce “GEMOR-GS: Global Extrinsic-calibration-refined Mapping and Online Relocalization for 3D Gaussian-Splatting SLAM” 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
Yuchen Yang, Ming Ding, Xinyu Wu, Hesheng Wang
GEMOR-GS builds a fast online map for tracking, then refines archived camera poses and reconstructs a 3D Gaussian splatting map offline using calibrated LiDAR–camera geometry. It estimates camera pose from new images against that map after tracking loss. Across six sequences, offline maps exceed online maps by 14.3–19.9 dB PSNR on 6,632 keyframes, while relocalization fails on weak-texture Bright (0/100).
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.