mlsanityData to 7 Oct 2026

Learning Instance-Aware Correspondences for Robust Multi-Instance Point Cloud Registration in Cluttered Scenes

arXiv 2404.04557 ↗First posted 6 Apr 2024cs.CVcited by 7

Not on the map. Why?

Abstract (arXiv)

Multi-instance point cloud registration estimates the poses of multiple instances of a model point cloud in a scene point cloud. Extracting accurate point correspondence is to the center of the problem. Existing approaches usually treat the scene point cloud as a whole, overlooking the separation of instances. Therefore, point features could be easily polluted by other points from the background or different instances, leading to inaccurate correspondences oblivious to separate instances, especially in cluttered scenes. In this work, we propose MIRETR, Multi-Instance REgistration TRansformer, a coarse-to-fine approach to the extraction of instance-aware correspondences. At the coarse level, it jointly learns instance-aware superpoint features and predicts per-instance masks. With instance masks, the influence from outside of the instance being concerned is minimized, such that highly reliable superpoint correspondences can be extracted. The superpoint correspondences are then extended to instance candidates at the fine level according to the instance masks. At last, an efficient candidate selection and refinement algorithm is devised to obtain the final registrations. Extensive experiments on three public benchmarks demonstrate the efficacy of our approach. In particular, MIRETR outperforms the state of the arts by 16.6 points on F1 score on the challenging ROBI benchmark. Code and models are available at https://github.com/zhiyuanYU134/MIRETR.

Key facts. Cited by 7 papers in the map (data through 7 Oct 2026). Builds on 45 of its 59 references that are in the map. First posted on arXiv 6 Apr 2024.

Builds on

45 of 59 references.

References · 45

  1. 2017Attention Is All You Needcited by 59,932
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  3. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,088
  4. 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,880
  5. 2016V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentationcited by 2,284
  6. 2019SuperGlue: Learning Feature Matching with Graph Neural Networkscited by 1,437
  7. 2019KPConv: Flexible and Deformable Convolution for Point Cloudscited by 1,202
  8. 2017PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenescited by 768
  9. 2019Normalized Object Coordinate Space for Category-Level 6D Object Pose and Size Estimationcited by 490
  10. 2019Deep Closest Point: Learning Representations for Point Cloud Registrationcited by 318
  11. 2020PREDATOR: Registration of 3D Point Clouds with Low Overlapcited by 287
  12. 2018BOP: Benchmark for 6D Object Pose Estimationcited by 254
33 more
  1. 2019PointNetLK: Robust & Efficient Point Cloud Registration using PointNetcited by 231
  2. 2018SO-Net: Self-Organizing Network for Point Cloud Analysiscited by 210
  3. 2022Geometric Transformer for Fast and Robust Point Cloud Registrationcited by 200
  4. 2020RPM-Net: Robust Point Matching using Learned Featurescited by 181
  5. 2018PPFNet: Global Context Aware Local Features for Robust 3D Point Matchingcited by 180
  6. 2020D3Feat: Joint Learning of Dense Detection and Description of 3D Local Featurescited by 175
  7. 2019Implicit 3D Orientation Learning for 6D Object Detection from RGB Imagescited by 170
  8. 2018The Perfect Match: 3D Point Cloud Matching with Smoothed Densitiescited by 169
  9. 2021CoFiNet: Reliable Coarse-to-fine Correspondences for Robust Point Cloud Registrationcited by 151
  10. 2019PRNet: Self-Supervised Learning for Partial-to-Partial Registrationcited by 146
  11. 2018Scan2CAD: Learning CAD Model Alignment in RGB-D Scanscited by 142
  12. 2021PointDSC: Robust Point Cloud Registration using Deep Spatial Consistencycited by 142
  13. 2018PPF-FoldNet: Unsupervised Learning of Rotation Invariant 3D Local Descriptorscited by 140
  14. 2020Learning Canonical Shape Space for Category-Level 6D Object Pose and Size Estimationcited by 123
  15. 2022ZebraPose: Coarse to Fine Surface Encoding for 6DoF Object Pose Estimationcited by 115
  16. 2020Feature-metric Registration: A Fast Semi-supervised Approach for Robust Point Cloud Registration without Correspondencescited by 108
  17. 2021DualPoseNet: Category-level 6D Object Pose and Size Estimation Using Dual Pose Network with Refined Learning of Pose Consistencycited by 103
  18. 2021Robust Point Cloud Registration Framework Based on Deep Graph Matchingcited by 96
  19. 2019Single-Stage 6D Object Pose Estimationcited by 88
  20. 2022Category-Level 6D Object Pose and Size Estimation using Self-Supervised Deep Prior Deformation Networkscited by 73
  21. 2018Holistic 3D Scene Parsing and Reconstruction from a Single RGB Imagecited by 69
  22. 2021ROCA: Robust CAD Model Retrieval and Alignment from a Single Imagecited by 66
  23. 2020Mask2CAD: 3D Shape Prediction by Learning to Segment and Retrievecited by 65
  24. 2021OMNet: Learning Overlapping Mask for Partial-to-Partial Point Cloud Registrationcited by 50
  25. 2020SceneCAD: Predicting Object Alignments and Layouts in RGB-D Scanscited by 44
  26. 2019Progressive-X: Efficient, Anytime, Multi-Model Fitting Algorithmcited by 43
  27. 2020CONSAC: Robust Multi-Model Fitting by Conditional Sample Consensuscited by 30
  28. 2021ROBI: A Multi-View Dataset for Reflective Objects in Robotic Bin-Pickingcited by 22
  29. 2022Sim-to-Real 6D Object Pose Estimation via Iterative Self-training for Robotic Bin Pickingcited by 20
  30. 2021StablePose: Learning 6D Object Poses from Geometrically Stable Patchescited by 17
  31. 2022Focal Length and Object Pose Estimation via Render and Comparecited by 17
  32. 2021Multi-instance Point Cloud Registration by Efficient Correspondence Clusteringcited by 7
  33. 2022PointCLM: A Contrastive Learning-based Framework for Multi-instance Point Cloud Registrationcited by 6

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 1; 2025: 5; 2026: 1.

7 papers.

  1. 2025BUFFER-X: Towards Zero-Shot Point Cloud Registration in Diverse Scenescited by 8
  2. 2025Progressive Correspondence Regenerator for Robust 3D Registrationcited by 5
  3. 20243D Registration in 30 Years: A Surveycited by 2
  4. 2025ZeroBP: Learning Position-Aware Correspondence for Zero-shot 6D Pose Estimation in Bin-Pickingcited by 2
  5. 2025Cross-modal feature fusion for robust point cloud registration with ambiguous geometrycited by 1
  6. 2026Towards Zero-Shot Point Cloud Registration Across Diverse Scales, Scenes, and Sensor Setupscited by 1
  7. 2025DINOReg: Strong Point Cloud Registration with Vision Foundation Modelcited by 0