mlsanityData to 7 Oct 2026

PointDSC: Robust Point Cloud Registration using Deep Spatial Consistency

arXiv 2103.05465 ↗First posted 9 Mar 2021cs.CVcited by 142

Not on the map. Why?

Abstract (arXiv)

Removing outlier correspondences is one of the critical steps for successful feature-based point cloud registration. Despite the increasing popularity of introducing deep learning methods in this field, spatial consistency, which is essentially established by a Euclidean transformation between point clouds, has received almost no individual attention in existing learning frameworks. In this paper, we present PointDSC, a novel deep neural network that explicitly incorporates spatial consistency for pruning outlier correspondences. First, we propose a nonlocal feature aggregation module, weighted by both feature and spatial coherence, for feature embedding of the input correspondences. Second, we formulate a differentiable spectral matching module, supervised by pairwise spatial compatibility, to estimate the inlier confidence of each correspondence from the embedded features. With modest computation cost, our method outperforms the state-of-the-art hand-crafted and learning-based outlier rejection approaches on several real-world datasets by a significant margin. We also show its wide applicability by combining PointDSC with different 3D local descriptors.

Key facts. Cited by 142 papers in the map (data through 7 Oct 2026). Builds on 43 of its 85 references that are in the map. First posted on arXiv 9 Mar 2021.

Builds on

43 of 85 references.

References · 43

  1. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  2. 2017Non-local Neural Networkscited by 2,290
  3. 2019SuperGlue: Learning Feature Matching with Graph Neural Networkscited by 1,437
  4. 20194D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networkscited by 1,201
  5. 2018Open3D: A Modern Library for 3D Data Processingcited by 717
  6. 2018PVNet: Pixel-wise Voting Network for 6DoF Pose Estimationcited by 385
  7. 20163DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructionscited by 333
  8. 2019Deep Closest Point: Learning Representations for Point Cloud Registrationcited by 318
  9. 2020PREDATOR: Registration of 3D Point Clouds with Low Overlapcited by 287
  10. 2020TEASER: Fast and Certifiable Point Cloud Registrationcited by 274
  11. 2019PointNetLK: Robust & Efficient Point Cloud Registration using PointNetcited by 231
  12. 2020Deep Global Registrationcited by 228
31 more
  1. 2019PVN3D: A Deep Point-wise 3D Keypoints Voting Network for 6DoF Pose Estimationcited by 204
  2. 2020RPM-Net: Robust Point Matching using Learned Featurescited by 181
  3. 2018PPFNet: Global Context Aware Local Features for Robust 3D Point Matchingcited by 180
  4. 2017Learning to Find Good Correspondencescited by 176
  5. 2020D3Feat: Joint Learning of Dense Detection and Description of 3D Local Featurescited by 175
  6. 2018The Perfect Match: 3D Point Cloud Matching with Smoothed Densitiescited by 169
  7. 2019PRNet: Self-Supervised Learning for Partial-to-Partial Registrationcited by 146
  8. 2020SpinNet: Learning a General Surface Descriptor for 3D Point Cloud Registrationcited by 141
  9. 2018PPF-FoldNet: Unsupervised Learning of Rotation Invariant 3D Local Descriptorscited by 140
  10. 2017Graph-Cut RANSACcited by 133
  11. 20183DFeat-Net: Weakly Supervised Local 3D Features for Point Cloud Registrationcited by 100
  12. 20193DRegNet: A Deep Neural Network for 3D Point Registrationcited by 96
  13. 2019ContextDesc: Local Descriptor Augmentation with Cross-Modality Contextcited by 79
  14. 2020Learning multiview 3D point cloud registrationcited by 78
  15. 2019USIP: Unsupervised Stable Interest Point Detection from 3D Point Cloudscited by 70
  16. 2017Guaranteed Outlier Removal for Point Cloud Registration with Correspondencescited by 62
  17. 2018GeoDesc: Learning Local Descriptors by Integrating Geometry Constraintscited by 52
  18. 2020End-to-End Learning Local Multi-view Descriptors for 3D Point Cloudscited by 47
  19. 2019A Polynomial-time Solution for Robust Registration with Extreme Outlier Ratescited by 38
  20. 2020ROBIN: a Graph-Theoretic Approach to Reject Outliers in Robust Estimation using Invariantscited by 38
  21. 2019NM-Net: Mining Reliable Neighbors for Robust Feature Correspondencescited by 29
  22. 2019SDRSAC: Semidefinite-Based Randomized Approach for Robust Point Cloud Registration without Correspondencescited by 26
  23. 20193D Local Features for Direct Pairwise Registrationcited by 26
  24. 2018Extreme Relative Pose Estimation for RGB-D Scans via Scene Completioncited by 22
  25. 2020Distinctive 3D local deep descriptorscited by 21
  26. 2019A Practical Maximum Clique Algorithm for Matching with Pairwise Constraintscited by 20
  27. 2020AdaLAM: Revisiting Handcrafted Outlier Detectioncited by 14
  28. 2019A Performance Evaluation of Correspondence Grouping Methods for 3D Rigid Data Matchingcited by 10
  29. 2017In search of inliers: 3d correspondence by local and global votingcited by 7
  30. 2018Learning and Matching Multi-View Descriptors for Registration of Point Cloudscited by 6
  31. 2018Performance Evaluation of 3D Correspondence Grouping Algorithmscited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 11; 2022: 23; 2023: 37; 2024: 32; 2025: 25; 2026: 14.

142 papers.

Sort:

By area:

  1. 2022Geometric Transformer for Fast and Robust Point Cloud Registrationcited by 200
  2. 2021Learning to Match Features with Seeded Graph Matching Networkcited by 92
  3. 2021Lepard: Learning partial point cloud matching in rigid and deformable scenescited by 85
  4. 20233D Registration with Maximal Cliquescited by 83
  5. 2023GeoTransformer: Fast and Robust Point Cloud Registration with Geometric Transformercited by 83
  6. 2021You Only Hypothesize Once: Point Cloud Registration with Rotation-equivariant Descriptorscited by 68
  7. 2022SC^2-PCR: A Second Order Spatial Compatibility for Efficient and Robust Point Cloud Registrationcited by 62
  8. 2021Deep Hough Voting for Robust Global Registrationcited by 53
  9. 2025Any6D: Model-free 6D Pose Estimation of Novel Objectscited by 47
  10. 2022CorrI2P: Deep Image-to-Point Cloud Registration via Dense Correspondencecited by 43
  11. 2022Non-rigid Point Cloud Registration with Neural Deformation Pyramidcited by 41
  12. 2023RegFormer: An Efficient Projection-Aware Transformer Network for Large-Scale Point Cloud Registrationcited by 37
38 more
  1. 2023Robust Outlier Rejection for 3D Registration with Variational Bayescited by 31
  2. 2023Open-vocabulary object 6D pose estimationcited by 29
  3. 20232D3D-MATR: 2D-3D Matching Transformer for Detection-free Registration between Images and Point Cloudscited by 27
  4. 2023Unsupervised Deep Probabilistic Approach for Partial Point Cloud Registrationcited by 26
  5. 2023Point-TTA: Test-Time Adaptation for Point Cloud Registration Using Multitask Meta-Auxiliary Learningcited by 23
  6. 2022Spectral Geometric Verification: Re-Ranking Point Cloud Retrieval for Metric Localizationcited by 21
  7. 2023Understanding and Improving Features Learned in Deep Functional Mapscited by 20
  8. 2022Reliable Inlier Evaluation for Unsupervised Point Cloud Registrationcited by 19
  9. 2022ImLoveNet: Misaligned Image-supported Registration Network for Low-overlap Point Cloud Pairscited by 17
  10. 2023Robust Multiview Point Cloud Registration with Reliable Pose Graph Initialization and History Reweightingcited by 17
  11. 2024FastMAC: Stochastic Spectral Sampling of Correspondence Graphcited by 16
  12. 2024Scalable 3D Registration via Truncated Entry-wise Absolute Residualscited by 16
  13. 2022ARCS: Accurate Rotation and Correspondence Searchcited by 15
  14. 2024Extend Your Own Correspondences: Unsupervised Distant Point Cloud Registration by Progressive Distance Extensioncited by 15
  15. 2023Deep Graph-based Spatial Consistency for Robust Non-rigid Point Cloud Registrationcited by 14
  16. 2023PointMBF: A Multi-scale Bidirectional Fusion Network for Unsupervised RGB-D Point Cloud Registrationcited by 14
  17. 2022Objects Can Move: 3D Change Detection by Geometric Transformation Constistencycited by 13
  18. 2022SRFeat: Learning Locally Accurate and Globally Consistent Non-Rigid Shape Correspondencecited by 13
  19. 2023PCR-CG: Point Cloud Registration via Deep Explicit Color and Geometrycited by 13
  20. 2021DetarNet: Decoupling Translation and Rotation by Siamese Network for Point Cloud Registrationcited by 12
  21. 2022Improving RGB-D Point Cloud Registration by Learning Multi-scale Local Linear Transformationcited by 12
  22. 2023Multi-Body Neural Scene Flowcited by 12
  23. 2024UNOPose: Unseen Object Pose Estimation with an Unposed RGB-D Reference Imagecited by 12
  24. 2025TurboReg: TurboClique for Robust and Efficient Point Cloud Registrationcited by 12
  25. 2023Density-invariant Features for Distant Point Cloud Registrationcited by 11
  26. 2023G3Reg: Pyramid Graph-based Global Registration using Gaussian Ellipsoid Modelcited by 11
  27. 2022A New Outlier Removal Strategy Based on Reliability of Correspondence Graph for Fast Point Cloud Registrationcited by 10
  28. 2024High-resolution open-vocabulary object 6D pose estimationcited by 10
  29. 2023Learning Accurate Template Matching with Differentiable Coarse-to-Fine Correspondence Refinementcited by 9
  30. 2023Cross-source Point Cloud Registration: Challenges, Progress and Prospectscited by 9
  31. 2023SPEAL: Skeletal Prior Embedded Attention Learning for Cross-Source Point Cloud Registrationcited by 9
  32. 2026CMHANet: A Cross-Modal Hybrid Attention Network for Point Cloud Registrationcited by 9
  33. 2023VSFormer: Visual-Spatial Fusion Transformer for Correspondence Pruningcited by 8
  34. 2024Efficient and Robust Point Cloud Registration via Heuristics-guided Parameter Searchcited by 8
  35. 2025Diff2I2P: Differentiable Image-to-Point Cloud Registration with Diffusion Priorcited by 8
  36. 2025Register Any Point: Scaling 3D Point Cloud Registration by Flow Matchingcited by 8
  37. 2021WSDesc: Weakly Supervised 3D Local Descriptor Learning for Point Cloud Registrationcited by 7
  38. 2021Multi-instance Point Cloud Registration by Efficient Correspondence Clusteringcited by 7

50 most cited of 142 All 142 (JSON)