mlsanityData to 7 Oct 2026

WSDesc: Weakly Supervised 3D Local Descriptor Learning for Point Cloud Registration

arXiv 2108.02740 ↗First posted 5 Aug 2021cs.CV cs.GRcited by 7

Not on the map. Why?

Abstract (arXiv)

In this work, we present a novel method called WSDesc to learn 3D local descriptors in a weakly supervised manner for robust point cloud registration. Our work builds upon recent 3D CNN-based descriptor extractors, which leverage a voxel-based representation to parameterize local geometry of 3D points. Instead of using a predefined fixed-size local support in voxelization, we propose to learn the optimal support in a data-driven manner. To this end, we design a novel differentiable voxelization layer that can back-propagate the gradient to the support size optimization. To train the extracted descriptors, we propose a novel registration loss based on the deviation from rigidity of 3D transformations, and the loss is weakly supervised by the prior knowledge that the input point clouds have partial overlap, without requiring ground-truth alignment information. Through extensive experiments, we show that our learned descriptors yield superior performance on existing geometric registration benchmarks.

Key facts. Cited by 7 papers in the map (data through 7 Oct 2026). Builds on 53 of its 79 references that are in the map. First posted on arXiv 5 Aug 2021.

Builds on

53 of 79 references.

References · 53

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  3. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,384
  4. 2015FaceNet: A Unified Embedding for Face Recognition and Clusteringcited by 2,833
  5. 2019KPConv: Flexible and Deformable Convolution for Point Cloudscited by 1,202
  6. 20194D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networkscited by 1,201
  7. 2016Instance Normalization: The Missing Ingredient for Fast Stylizationcited by 1,187
  8. 2017In Defense of the Triplet Loss for Person Re-Identificationcited by 784
  9. 2018Open3D: A Modern Library for 3D Data Processingcited by 717
  10. 2018PointConv: Deep Convolutional Networks on 3D Point Cloudscited by 656
  11. 2020PointContrast: Unsupervised Pre-training for 3D Point Cloud Understandingcited by 502
  12. 2018Spherical CNNscited by 452
38 more
  1. 2019Soft Rasterizer: A Differentiable Renderer for Image-based 3D Reasoningcited by 420
  2. 20163DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructionscited by 333
  3. 2019Deep Closest Point: Learning Representations for Point Cloud Registrationcited by 318
  4. 2020PREDATOR: Registration of 3D Point Clouds with Low Overlapcited by 287
  5. 2017Learning SO(3) Equivariant Representations with Spherical CNNscited by 244
  6. 2016Volumetric and Multi-View CNNs for Object Classification on 3D Datacited by 240
  7. 2019PointNetLK: Robust & Efficient Point Cloud Registration using PointNetcited by 231
  8. 2020Deep Global Registrationcited by 228
  9. 2020RPM-Net: Robust Point Matching using Learned Featurescited by 181
  10. 2018PPFNet: Global Context Aware Local Features for Robust 3D Point Matchingcited by 180
  11. 2017Learning to Find Good Correspondencescited by 176
  12. 2020D3Feat: Joint Learning of Dense Detection and Description of 3D Local Featurescited by 175
  13. 2018The Perfect Match: 3D Point Cloud Matching with Smoothed Densitiescited by 169
  14. 2016Go-ICP: A Globally Optimal Solution to 3D ICP Point-Set Registrationcited by 159
  15. 2019PRNet: Self-Supervised Learning for Partial-to-Partial Registrationcited by 146
  16. 2021PointDSC: Robust Point Cloud Registration using Deep Spatial Consistencycited by 142
  17. 2020SpinNet: Learning a General Surface Descriptor for 3D Point Cloud Registrationcited by 141
  18. 2018PPF-FoldNet: Unsupervised Learning of Rotation Invariant 3D Local Descriptorscited by 140
  19. 2020Feature-metric Registration: A Fast Semi-supervised Approach for Robust Point Cloud Registration without Correspondencescited by 108
  20. 20183D Point Capsule Networkscited by 106
  21. 20183DFeat-Net: Weakly Supervised Local 3D Features for Point Cloud Registrationcited by 100
  22. 2018Neural Nearest Neighbors Networkscited by 98
  23. 2017PCPNET: Learning Local Shape Properties from Raw Point Cloudscited by 97
  24. 20193DRegNet: A Deep Neural Network for 3D Point Registrationcited by 96
  25. 2020DeepGMR: Learning Latent Gaussian Mixture Models for Registrationcited by 91
  26. 2018Unsupervised Deep Learning for Structured Shape Matchingcited by 81
  27. 2020Learning multiview 3D point cloud registrationcited by 78
  28. 2021OMNet: Learning Overlapping Mask for Partial-to-Partial Point Cloud Registrationcited by 50
  29. 2020End-to-End Learning Local Multi-view Descriptors for 3D Point Cloudscited by 47
  30. 2017Learning Compact Geometric Featurescited by 45
  31. 2020DH3D: Deep Hierarchical 3D Descriptors for Robust Large-Scale 6DoF Relocalizationcited by 42
  32. 2013Rotational Projection Statistics for 3D Local Surface Description and Object Recognitioncited by 36
  33. 20193D Local Features for Direct Pairwise Registrationcited by 26
  34. 2020Learning to Orient Surfaces by Self-supervised Spherical CNNscited by 26
  35. 2021Self-supervised Geometric Perceptioncited by 26
  36. 2020Distinctive 3D local deep descriptorscited by 21
  37. 2020Recurrent Multi-view Alignment Network for Unsupervised Surface Registrationcited by 19
  38. 2017Learning Local Shape Descriptors from Part Correspondences With Multi-view Convolutional Networkscited by 17

50 most cited of 53 All 53 (JSON)

Built on by

Cited by year

Citing papers per year: 2021: 1; 2022: 0; 2023: 3; 2024: 3.

7 papers.

  1. 2023FreeZe: Training-free zero-shot 6D pose estimation with geometric and vision foundation modelscited by 56
  2. 20213D Question Answeringcited by 41
  3. 2024Multiview Equivariance Improves 3D Correspondence Understanding with Minimal Feature Finetuningcited by 13
  4. 2023Generalizable Local Feature Pre-training for Deformable Shape Analysiscited by 6
  5. 2024Point-Level Topological Representation Learning on Point Cloudscited by 4
  6. 2023RoCNet: 3D Robust Registration of Point-Clouds using Deep Learningcited by 2
  7. 20243D Registration in 30 Years: A Surveycited by 2