… of point cloud registration and build a new cross-source point cloud benchmark to evaluate
the performance of the state-ofthe-art registration methods in solving these challenges. …

… 3-D registration in the realistic case where many point-topoint correspondences are
outliers due to incorrect matching. (a) Bunny dataset (source cloud in blue and target cloud in …

… think about representing point clouds. For classification and … of PointNet to point cloud
registration has remained elusive. … -of-the-art in common registration scenarios. The architecture …

… We begin with a new algorithm for learned pairwise point cloud registration, which uses
two point clouds as input and outputs estimated transformation parameters (Sec. 3.1). This …

… 3D point cloud registration framework that achieves comparable registration accuracy to …
of candidates, which can boost the registration accuracy. We comprehensively validate the ef…

… PointNetLK [2], this work introduces point cloud registration network (PCRNet), a framework
for estimating the misalignment between two point clouds … directly process point clouds for …

… Since we use per-point embeddings of the two input point clouds to generate a mapping m
… seek a feature per point in the input point clouds rather than one feature per cloud. For this …

… correspondences for point cloud registration. Recent keypoint… great potential in registration.
They seek correspondences over … the geometric structure of the point clouds. We propose …

… Extracting robust and general 3D local features is key to downstream tasks such as point
cloud registration and reconstruction. Existing learning-based local descriptors are either …

… In this paper, we focus on leveraging the spatial consistency in outlier rejection for robust
3D point cloud registration. Spectral matching (SM) [36] is a well-known traditional algorithm …