Registration and matching: research map
272 papers on Registration and matching in Point clouds, from ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR (2016–2026), grouped into 2 clusters and 4 approaches. The busiest year so far is 2025.
What it is
Registration aligns two point clouds of the same object or scene taken from different positions, by finding matching points and the transformation between them. It is a basic step in 3D mapping and reconstruction.
Terms you will see on this page:
- Point cloud registration
- estimating the rotation and translation that align two scans.
- Correspondence
- a pair of points believed to be the same physical point.
- Outlier rejection
- discarding wrong correspondences.
Start reading with the most cited papers here:
- PointNetLK: Robust & Efficient Point Cloud Registration Using PointNet (CVPR 2019 · 891 citations)
- Predator: Registration of 3D Point Clouds With Low Overlap (CVPR 2021 · 715 citations)
Within Point clouds, its share grew from 7.7% in 2023–24 to 9.0% in 2025–26 (35 → 46 papers at ICML, NeurIPS, ICLR, CVPR and AAAI, the venues with data for all four years).
Explore Registration and matching in the interactive map
Working on something in this topic? Describe your idea in scime atlas to see which approach it falls under, the closest papers by meaning and how crowded the spot has become.
What changed recently
Approaches whose share of the topic moved most between 2023–24 and 2025–26 (ICML, NeurIPS, ICLR, CVPR and AAAI):
- Learned feature-based registration rising: 54% → 63% of the topic (19 → 29 papers)
Papers from 2025–2026 that are already cited most:
- QuCOOP: A Versatile Framework for Solving Composite and Binary-Parametrised Problems on Quantum Annealers (CVPR 2025 · 3 citations)
- Gromov-Wasserstein Problem with Cyclic Symmetry (CVPR 2025 · 2 citations)
Approaches and key papers
Feature-based registration · 192 papers
Keywords: cloud registration, overlap, feature
Approaches in this cluster:
- Learned feature-based registration (125 papers)
Register point clouds with learned features, multi-view and image cues, and unsupervised training. - Robust inlier selection for registration (67 papers)
Reject outlier correspondences using geometric transformers, inlier evaluation and randomized solvers.
Most cited and most cited since 2024:
- PointNetLK: Robust & Efficient Point Cloud Registration Using PointNet (CVPR 2019 · 891 citations)
- Predator: Registration of 3D Point Clouds With Low Overlap (CVPR 2021 · 715 citations)
- FastMAC: Stochastic Spectral Sampling of Correspondence Graph (CVPR 2024 · 27 citations)
- Dynamic Cues-Assisted Transformer for Robust Point Cloud Registration (CVPR 2024 · 25 citations)
Rigid and non-rigid registration · 80 papers
Keywords: rigid, point registration, non
Approaches in this cluster:
- Non-rigid point set registration (67 papers)
Align point sets with rigid and non-rigid transformations using coherent drift, graph consistency and transformers. - Quantum and certifiable correspondence (13 papers)
Solve point-set correspondence and transformation estimation with quantum or certifiably robust solvers.
Most cited and most cited since 2024:
- RPM-Net: Robust Point Matching Using Learned Features (CVPR 2020 · 623 citations)
- 3DRegNet: A Deep Neural Network for 3D Point Registration (CVPR 2020 · 276 citations)
- Scalable 3D Registration via Truncated Entry-wise Absolute Residuals (CVPR 2024 · 12 citations)
- Multiway Point Cloud Mosaicking with Diffusion and Global Optimization (CVPR 2024 · 11 citations)
Related topics in Point clouds
- Point cloud classification (1,097)
- Segmentation and detection (567)
- Shape and surface completion (558)
