5 papers · 1 filter
TeaMatch: Teachable Cross-Modal Representation Learning for 2D-3D Matching
Chongjian Wang, Junjie Gao
Learning reliable correspondences between images and point clouds is fundamental for 2D-3D matching. Despite recent progress in detection-free methods, existing approaches primaril…
SHReg: Strictly Rotation-Equivariant Point Cloud Registration via Spherical Harmonics
Chongjian Wang, Junjie Gao
Point cloud registration critically depends on local features that are both distinctive and robust to arbitrary 3D rotations. Existing learning-based methods typically approximate…
Deep-PE: A Learning-Based Pose Evaluator for Point Cloud Registration
Junjie Gao, Chongjian Wang, Zhongjun Ding +4
In the realm of point cloud registration, the most prevalent pose evaluation approaches are statistics-based, identifying the optimal transformation by maximizing the number of con…
D3Former: Jointly Learning Repeatable Dense Detectors and Feature-enhanced Descriptors via Saliency-guided Transformer
Junjie Gao, Pengfei Wang, Qiujie Dong +3
Establishing accurate and representative matches is a crucial step in addressing the point cloud registration problem. A commonly employed approach involves detecting keypoints wit…
OAAFormer: Robust and Efficient Point Cloud Registration Through Overlapping-Aware Attention in Transformer
Junjie Gao, Qiujie Dong, Ruian Wang +4
In the domain of point cloud registration, the coarse-to-fine feature matching paradigm has received substantial attention owing to its impressive performance. This paradigm involv…