mlsanityData to 7 Oct 2026

3DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructions

arXiv 1603.08182 ↗First posted 27 Mar 2016cs.CVcited by 333

Not on the map. Why?

Abstract (arXiv)

Matching local geometric features on real-world depth images is a challenging task due to the noisy, low-resolution, and incomplete nature of 3D scan data. These difficulties limit the performance of current state-of-art methods, which are typically based on histograms over geometric properties. In this paper, we present 3DMatch, a data-driven model that learns a local volumetric patch descriptor for establishing correspondences between partial 3D data. To amass training data for our model, we propose a self-supervised feature learning method that leverages the millions of correspondence labels found in existing RGB-D reconstructions. Experiments show that our descriptor is not only able to match local geometry in new scenes for reconstruction, but also generalize to different tasks and spatial scales (e.g. instance-level object model alignment for the Amazon Picking Challenge, and mesh surface correspondence). Results show that 3DMatch consistently outperforms other state-of-the-art approaches by a significant margin. Code, data, benchmarks, and pre-trained models are available online at http://3dmatch.cs.princeton.edu

Key facts. Cited by 333 papers in the map (data through 7 Oct 2026). First posted on arXiv 27 Mar 2016.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 51; 2022: 59; 2023: 68; 2024: 65; 2025: 63; 2026: 27.

333 papers.

Sort:

By area:

  1. 2021Depth-supervised NeRF: Fewer Views and Faster Training for Freecited by 732
  2. 2023DL3DV-10K: A Large-Scale Scene Dataset for Deep Learning-based 3D Visioncited by 514
  3. 2021MonoScene: Monocular 3D Semantic Scene Completioncited by 357
  4. 2022Geometric Transformer for Fast and Robust Point Cloud Registrationcited by 200
  5. 2021The MVTec 3D-AD Dataset for Unsupervised 3D Anomaly Detection and Localizationcited by 174
  6. 2021CoFiNet: Reliable Coarse-to-fine Correspondences for Robust Point Cloud Registrationcited by 151
  7. 2023BundleSDF: Neural 6-DoF Tracking and 3D Reconstruction of Unknown Objectscited by 144
  8. 2021PointDSC: Robust Point Cloud Registration using Deep Spatial Consistencycited by 142
  9. 2025Video World Models with Long-term Spatial Memorycited by 124
  10. 2021Robust Point Cloud Registration Framework Based on Deep Graph Matchingcited by 96
  11. 2022REGTR: End-to-end Point Cloud Correspondences with Transformerscited by 94
  12. 2022Back to the Feature: Classical 3D Features are (Almost) All You Need for 3D Anomaly Detectioncited by 92
38 more
  1. 2021Lepard: Learning partial point cloud matching in rigid and deformable scenescited by 85
  2. 2021A comprehensive survey on point cloud registrationcited by 84
  3. 20233D Registration with Maximal Cliquescited by 83
  4. 2023GeoTransformer: Fast and Robust Point Cloud Registration with Geometric Transformercited by 83
  5. 2021Equivariant Point Network for 3D Point Cloud Analysiscited by 81
  6. 2024Explore until Confident: Efficient Exploration for Embodied Question Answeringcited by 74
  7. 2024How NeRFs and 3D Gaussian Splatting are Reshaping SLAM: a Surveycited by 72
  8. 2021You Only Hypothesize Once: Point Cloud Registration with Rotation-equivariant Descriptorscited by 68
  9. 2023Rotation-Invariant Transformer for Point Cloud Matchingcited by 68
  10. 2022Anomaly Detection in 3D Point Clouds using Deep Geometric Descriptorscited by 66
  11. 2024GS2Mesh: Surface Reconstruction from Gaussian Splatting via Novel Stereo Viewscited by 65
  12. 2022SC^2-PCR: A Second Order Spatial Compatibility for Efficient and Robust Point Cloud Registrationcited by 62
  13. 2022A Survey of Non-Rigid 3D Registrationcited by 59
  14. 2024LoopSplat: Loop Closure by Registering 3D Gaussian Splatscited by 59
  15. 2021DeepI2P: Image-to-Point Cloud Registration via Deep Classificationcited by 58
  16. 2021DPFM: Deep Partial Functional Mapscited by 58
  17. 2023FreeZe: Training-free zero-shot 6D pose estimation with geometric and vision foundation modelscited by 56
  18. 2022Unsupervised Point Cloud Representation Learning with Deep Neural Networks: A Surveycited by 54
  19. 2022NeurAR: Neural Uncertainty for Autonomous 3D Reconstruction with Implicit Neural Representationscited by 54
  20. 2021Deep Hough Voting for Robust Global Registrationcited by 53
  21. 2021OMNet: Learning Overlapping Mask for Partial-to-Partial Point Cloud Registrationcited by 50
  22. 2022SceneRF: Self-Supervised Monocular 3D Scene Reconstruction with Radiance Fieldscited by 49
  23. 2021Point-set Distances for Learning Representations of 3D Point Cloudscited by 48
  24. 2021UnsupervisedR&R: Unsupervised Point Cloud Registration via Differentiable Renderingcited by 44
  25. 2024Human-3Diffusion: Realistic Avatar Creation via Explicit 3D Consistent Diffusion Modelscited by 44
  26. 2022CorrI2P: Deep Image-to-Point Cloud Registration via Dense Correspondencecited by 43
  27. 2021P2-Net: Joint Description and Detection of Local Features for Pixel and Point Matchingcited by 42
  28. 2021Accurate Point Cloud Registration with Robust Optimal Transportcited by 38
  29. 2021RandomRooms: Unsupervised Pre-training from Synthetic Shapes and Randomized Layouts for 3D Object Detectioncited by 37
  30. 2023RegFormer: An Efficient Projection-Aware Transformer Network for Large-Scale Point Cloud Registrationcited by 37
  31. 2021AdaFusion: Visual-LiDAR Fusion with Adaptive Weights for Place Recognitioncited by 36
  32. 2023SCADE: NeRFs from Space Carving with Ambiguity-Aware Depth Estimatescited by 35
  33. 2023Differentiable Registration of Images and LiDAR Point Clouds with VoxelPoint-to-Pixel Matchingcited by 35
  34. 2022Transformers in 3D Point Clouds: A Surveycited by 31
  35. 2023Robust Outlier Rejection for 3D Registration with Variational Bayescited by 31
  36. 2022Generalized Differentiable RANSACcited by 30
  37. 2023SGAligner : 3D Scene Alignment with Scene Graphscited by 30
  38. 2021Robust and Accurate Superquadric Recovery: a Probabilistic Approachcited by 29

50 most cited of 333 All 333 (JSON)