Reproduce “SparseGraph-I2P: Consistency-Guided Hypothesis Selection for Efficient Image-to-Point-Cloud Registration” 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

Ziyang Yu, Xiang Li, Qiong Chang, Jun Miyazaki

SparseGraph-I2P addresses image-to-point-cloud camera relocalization by ranking candidate patch pairs using depth-derived geometric agreement, then reusing fine comparisons and filtering correspondences before pose estimation. It builds on a frozen 2D3D-MATR backbone. On RGB-D Scenes V2, it reaches 85.6% mean registration recall and 33.4 ms latency, but Stairs on 7-Scenes is an exception to its benchmark gains.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.