Search Results for author:
Found 2 papers, 0 papers with code
SCENES: Subpixel Correspondence Estimation With Epipolar Supervision
We relax this assumption by removing the requirement of 3D structure, e. g., depth maps or point clouds, and only require camera pose information, which can be obtained from odometry.
LoCUS: Learning Multiscale 3D-consistent Features from Posed Images
We start from the idea that the training objective can be framed as a patch retrieval problem: given an image patch in one view of a scene, we would like to retrieve (with high precision and recall) all patches in other views that map to the same real-world location.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
