mlsanityData to 7 Oct 2026

Learning Compact Geometric Features

arXiv 1709.05056 ↗First posted 15 Sep 2017cs.CV cs.GR cs.LGcited by 45

Not on the map. Why?

Abstract (arXiv)

We present an approach to learning features that represent the local geometry around a point in an unstructured point cloud. Such features play a central role in geometric registration, which supports diverse applications in robotics and 3D vision. Current state-of-the-art local features for unstructured point clouds have been manually crafted and none combines the desirable properties of precision, compactness, and robustness. We show that features with these properties can be learned from data, by optimizing deep networks that map high-dimensional histograms into low-dimensional Euclidean spaces. The presented approach yields a family of features, parameterized by dimension, that are both more compact and more accurate than existing descriptors.

Key facts. Cited by 45 papers in the map (data through 7 Oct 2026). First posted on arXiv 15 Sep 2017.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 18; 2022: 9; 2023: 6; 2024: 6; 2025: 4; 2026: 2.

45 papers.

Sort:

  1. 2022REGTR: End-to-end Point Cloud Correspondences with Transformerscited by 94
  2. 2021Lepard: Learning partial point cloud matching in rigid and deformable scenescited by 85
  3. 2021A comprehensive survey on point cloud registrationcited by 84
  4. 2021Equivariant Point Network for 3D Point Cloud Analysiscited by 81
  5. 2022CorrI2P: Deep Image-to-Point Cloud Registration via Dense Correspondencecited by 43
  6. 2021P2-Net: Joint Description and Detection of Local Features for Pixel and Point Matchingcited by 42
  7. 2021Learning general and distinctive 3D local deep descriptors for point cloud registrationcited by 41
  8. 2021Bootstrap Your Own Correspondencescited by 27
  9. 2021Self-supervised Geometric Perceptioncited by 26
  10. 2022Bending Graphs: Hierarchical Shape Matching using Gated Optimal Transportcited by 24
  11. 2021PCAM: Product of Cross-Attention Matrices for Rigid Registration of Point Cloudscited by 20
  12. 2022E2PN: Efficient SE(3)-Equivariant Point Networkcited by 17
33 more
  1. 2021PRIN/SPRIN: On Extracting Point-wise Rotation Invariant Featurescited by 15
  2. 2024Multiway Point Cloud Mosaicking with Diffusion and Global Optimizationcited by 14
  3. 2024MUFASA: Multi-View Fusion and Adaptation Network with Spatial Awareness for Radar Object Detectioncited by 13
  4. 2021IMFNet: Interpretable Multimodal Fusion for Point Cloud Registrationcited by 12
  5. 2021Exploiting Local Geometry for Feature and Graph Construction for Better 3D Point Cloud Processing with Graph Neural Networkscited by 9
  6. 2021RPSRNet: End-to-End Trainable Rigid Point Set Registration Network using Barnes-Hut 2D-Tree Representationcited by 8
  7. 2021City-scale Scene Change Detection using Point Cloudscited by 7
  8. 2021Planning with Learned Dynamic Model for Unsupervised Point Cloud Registrationcited by 7
  9. 2021WSDesc: Weakly Supervised 3D Local Descriptor Learning for Point Cloud Registrationcited by 7
  10. 2022CodedVTR: Codebook-based Sparse Voxel Transformer with Geometric Guidancecited by 7
  11. 2023Generalizable Local Feature Pre-training for Deformable Shape Analysiscited by 6
  12. 2023CRIN: Rotation-Invariant Point Cloud Analysis and Rotation Estimation via Centrifugal Reference Framecited by 4
  13. 2023KDD-LOAM: Jointly Learned Keypoint Detector and Descriptors Assisted LiDAR Odometry and Mappingcited by 4
  14. 2022Learning a Task-specific Descriptor for Robust Matching of 3D Point Cloudscited by 3
  15. 2021(Just) A Spoonful of Refinements Helps the Registration Error Go Downcited by 2
  16. 2021Understanding Pixel-level 2D Image Semantics with 3D Keypoint Knowledge Enginecited by 2
  17. 2023RoCNet: 3D Robust Registration of Point-Clouds using Deep Learningcited by 2
  18. 2024L-PR: Exploiting LiDAR Fiducial Marker for Unordered Low Overlap Multiview Point Cloud Registrationcited by 2
  19. 2024Towards Energy-Efficiency by Navigating the Trilemma of Energy, Latency, and Accuracycited by 2
  20. 20243D Registration in 30 Years: A Surveycited by 2
  21. 2025CanonNet: Canonical Ordering and Curvature Learning for Point Cloud Analysiscited by 2
  22. 2025Modality-Aware Feature Matching in Visual and Vision-Language Applications: A Comprehensive Surveycited by 2
  23. 2026RegFormer++: An Efficient Large-Scale 3D LiDAR Point Registration Network with Projection-Aware 2D Transformercited by 2
  24. 2022The Z-axis, X-axis, Weight and Disambiguation Methods for Constructing Local Reference Frame in 3D Registration: An Evaluationcited by 1
  25. 2022Learning Dense Features for Point Cloud Registration Using a Graph Attention Networkcited by 1
  26. 20233D Keypoint Estimation Using Implicit Representation Learningcited by 1
  27. 2023A Strong Baseline for Point Cloud Registration via Direct Superpoints Matchingcited by 1
  28. 2021What Stops Learning-based 3D Registration from Working in the Real World?cited by 0
  29. 2022Challenging the Universal Representation of Deep Models for 3D Point Cloud Registrationcited by 0
  30. 2024Ultron: Enabling Temporal Geometry Compression of 3D Mesh Sequences using Temporal Correspondence and Mesh Deformationcited by 0
  31. 2025DualReg: Dual-Space Filtering and Reinforcement for Rigid Registrationcited by 0
  32. 2025Attention-guided reference point shifting for Gaussian-mixture-based partial point set registrationcited by 0
  33. 2026One Shot Learning for Edge Detection on Point Cloudscited by 0