mlsanityData to 14 Oct 2026

PointRegGPT: Boosting 3D Point Cloud Registration using Generative Point-Cloud Pairs for Training

arXiv 2407.14054 ↗First posted 19 Jul 2024cs.CVcited by 4

Not on the map. Why?

Abstract (arXiv)

Data plays a crucial role in training learning-based methods for 3D point cloud registration. However, the real-world dataset is expensive to build, while rendering-based synthetic data suffers from domain gaps. In this work, we present PointRegGPT, boosting 3D point cloud registration using generative point-cloud pairs for training. Given a single depth map, we first apply a random camera motion to re-project it into a target depth map. Converting them to point clouds gives a training pair. To enhance the data realism, we formulate a generative model as a depth inpainting diffusion to process the target depth map with the re-projected source depth map as the condition. Also, we design a depth correction module to alleviate artifacts caused by point penetration during the re-projection. To our knowledge, this is the first generative approach that explores realistic data generation for indoor point cloud registration. When equipped with our approach, several recent algorithms can improve their performance significantly and achieve SOTA consistently on two common benchmarks. The code and dataset will be released on https://github.com/Chen-Suyi/PointRegGPT.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Builds on 49 of its 66 references that are in the map. First posted on arXiv 19 Jul 2024.

Builds on

49 of 66 references.

References · 49

  1. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  3. 2020Denoising Diffusion Implicit Modelscited by 9,781
  4. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,255
  5. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,058
  6. 2022Classifier-Free Diffusion Guidancecited by 5,847
  7. 2022Elucidating the Design Space of Diffusion-Based Generative Modelscited by 3,045
  8. 2021Image Super-Resolution via Iterative Refinementcited by 1,511
  9. 2023Zero-1-to-3: Zero-shot One Image to 3D Objectcited by 1,489
  10. 2022Human Motion Diffusion Modelcited by 1,080
  11. 2022Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Modelcited by 583
  12. 2021Solving Inverse Problems in Medical Imaging with Score-Based Generative Modelscited by 521
37 more
  1. 20163DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructionscited by 336
  2. 2019Deep Closest Point: Learning Representations for Point Cloud Registrationcited by 321
  3. 2020PREDATOR: Registration of 3D Point Clouds with Low Overlapcited by 290
  4. 2019PointNetLK: Robust & Efficient Point Cloud Registration using PointNetcited by 233
  5. 2020Deep Global Registrationcited by 229
  6. 2021More Control for Free! Image Synthesis with Semantic Diffusion Guidancecited by 212
  7. 2022Geometric Transformer for Fast and Robust Point Cloud Registrationcited by 202
  8. 2022Scalable Adaptive Computation for Iterative Generationcited by 195
  9. 2023Image Restoration with Mean-Reverting Stochastic Differential Equationscited by 183
  10. 2020RPM-Net: Robust Point Matching using Learned Featurescited by 182
  11. 2020D3Feat: Joint Learning of Dense Detection and Description of 3D Local Featurescited by 177
  12. 2018The Perfect Match: 3D Point Cloud Matching with Smoothed Densitiescited by 169
  13. 2021CoFiNet: Reliable Coarse-to-fine Correspondences for Robust Point Cloud Registrationcited by 152
  14. 2019PRNet: Self-Supervised Learning for Partial-to-Partial Registrationcited by 146
  15. 2020SpinNet: Learning a General Surface Descriptor for 3D Point Cloud Registrationcited by 142
  16. 2021PointDSC: Robust Point Cloud Registration using Deep Spatial Consistencycited by 142
  17. 2020Feature-metric Registration: A Fast Semi-supervised Approach for Robust Point Cloud Registration without Correspondencescited by 109
  18. 2021Robust Point Cloud Registration Framework Based on Deep Graph Matchingcited by 97
  19. 20193DRegNet: A Deep Neural Network for 3D Point Registrationcited by 96
  20. 2022REGTR: End-to-end Point Cloud Correspondences with Transformerscited by 96
  21. 2021Lepard: Learning partial point cloud matching in rigid and deformable scenescited by 85
  22. 20233D Registration with Maximal Cliquescited by 84
  23. 2021You Only Hypothesize Once: Point Cloud Registration with Rotation-equivariant Descriptorscited by 68
  24. 2021OMNet: Learning Overlapping Mask for Partial-to-Partial Point Cloud Registrationcited by 51
  25. 2022RGBD2: Generative Scene Synthesis via Incremental View Inpainting using RGBD Diffusion Modelscited by 45
  26. 2021Learning general and distinctive 3D local deep descriptors for point cloud registrationcited by 42
  27. 2023RegFormer: An Efficient Projection-Aware Transformer Network for Large-Scale Point Cloud Registrationcited by 37
  28. 2023Robust Outlier Rejection for 3D Registration with Variational Bayescited by 31
  29. 2023Point-TTA: Test-Time Adaptation for Point Cloud Registration Using Multitask Meta-Auxiliary Learningcited by 23
  30. 2021Moving in a 360 World: Synthesizing Panoramic Parallaxes from a Single Panoramacited by 19
  31. 2021FINet: Dual Branches Feature Interaction for Partial-to-Partial Point Cloud Registrationcited by 14
  32. 2024RecDiffusion: Rectangling for Image Stitching with Diffusion Modelscited by 14
  33. 2024HandBooster: Boosting 3D Hand-Mesh Reconstruction by Conditional Synthesis and Sampling of Hand-Object Interactionscited by 13
  34. 2023Density-invariant Features for Distant Point Cloud Registrationcited by 12
  35. 20213D Point Cloud Registration with Multi-Scale Architecture and Unsupervised Transfer Learningcited by 11
  36. 2024Single Image Rolling Shutter Removal with Diffusion Modelscited by 5
  37. 2023AutoSynth: Learning to Generate 3D Training Data for Object Point Cloud Registrationcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 1; 2025: 2; 2026: 1.

4 papers.

  1. 2025Unposed 3DGS Reconstruction with Probabilistic Procrustes Mappingcited by 8
  2. 20243D Registration in 30 Years: A Surveycited by 2
  3. 2025HybridReg: Robust 3D Point Cloud Registration with Hybrid Motionscited by 2
  4. 2026Robust 3D Alignment of Generative Reconstructions via Partial Monocular Observationscited by 0