mlsanityData to 14 Oct 2026

Leveraging Cycle-Consistent Anchor Points for Self-Supervised RGB-D Registration

arXiv 2510.14354 ↗First posted 16 Oct 2025cs.CV cs.ROcited by 0

Not on the map. Why?

Abstract (arXiv)

With the rise in consumer depth cameras, a wealth of unlabeled RGB-D data has become available. This prompts the question of how to utilize this data for geometric reasoning of scenes. While many RGB-D registration meth- ods rely on geometric and feature-based similarity, we take a different approach. We use cycle-consistent keypoints as salient points to enforce spatial coherence constraints during matching, improving correspondence accuracy. Additionally, we introduce a novel pose block that combines a GRU recurrent unit with transformation synchronization, blending historical and multi-view data. Our approach surpasses previous self- supervised registration methods on ScanNet and 3DMatch, even outperforming some older supervised methods. We also integrate our components into existing methods, showing their effectiveness.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 38 of its 48 references that are in the map. First posted on arXiv 16 Oct 2025.

Builds on

38 of 48 references.

References · 38

  1. 2017Attention Is All You Needcited by 60,361
  2. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  3. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  4. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,117
  5. 2019SuperGlue: Learning Feature Matching with Graph Neural Networkscited by 1,443
  6. 2018On the Continuity of Rotation Representations in Neural Networkscited by 1,214
  7. 2021LoFTR: Detector-Free Local Feature Matching with Transformerscited by 984
  8. 2021DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Camerascited by 687
  9. 20163DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructionscited by 336
  10. 2019Deep Closest Point: Learning Representations for Point Cloud Registrationcited by 321
  11. 2020PREDATOR: Registration of 3D Point Clouds with Low Overlapcited by 290
  12. 2020Space-Time Correspondence as a Contrastive Random Walkcited by 239
26 more
  1. 2020Deep Global Registrationcited by 229
  2. 2021Rethinking and Improving Relative Position Encoding for Vision Transformercited by 207
  3. 2022Geometric Transformer for Fast and Robust Point Cloud Registrationcited by 202
  4. 2020RPM-Net: Robust Point Matching using Learned Featurescited by 182
  5. 2020D3Feat: Joint Learning of Dense Detection and Description of 3D Local Featurescited by 177
  6. 2018The Perfect Match: 3D Point Cloud Matching with Smoothed Densitiescited by 169
  7. 2021PointDSC: Robust Point Cloud Registration using Deep Spatial Consistencycited by 142
  8. 2018PPF-FoldNet: Unsupervised Learning of Rotation Invariant 3D Local Descriptorscited by 141
  9. 2020Feature-metric Registration: A Fast Semi-supervised Approach for Robust Point Cloud Registration without Correspondencescited by 109
  10. 2020Learning Feature Descriptors using Camera Pose Supervisioncited by 104
  11. 2020DeepGMR: Learning Latent Gaussian Mixture Models for Registrationcited by 91
  12. 2020Learning multiview 3D point cloud registrationcited by 78
  13. 2019DeepICP: An End-to-End Deep Neural Network for 3D Point Cloud Registrationcited by 77
  14. 2022Theseus: A Library for Differentiable Nonlinear Optimizationcited by 65
  15. 2021UnsupervisedR&R: Unsupervised Point Cloud Registration via Differentiable Renderingcited by 44
  16. 2022Learning Pixel Trajectories with Multiscale Contrastive Random Walkscited by 36
  17. 2020PointGMM: a Neural GMM Network for Point Cloudscited by 33
  18. 2021Bootstrap Your Own Correspondencescited by 27
  19. 2023Unsupervised Deep Probabilistic Approach for Partial Point Cloud Registrationcited by 27
  20. 2019SDRSAC: Semidefinite-Based Randomized Approach for Robust Point Cloud Registration without Correspondencescited by 26
  21. 2022Reliable Inlier Evaluation for Unsupervised Point Cloud Registrationcited by 20
  22. 2020Unsupervised Point Cloud Registration via Salient Points Analysis (SPA)cited by 19
  23. 2018Synchronisation of Partial Multi-Matchings via Non-negative Factorisationscited by 15
  24. 2023PCR-CG: Point Cloud Registration via Deep Explicit Color and Geometrycited by 13
  25. 2022Improving RGB-D Point Cloud Registration by Learning Multi-scale Local Linear Transformationcited by 12
  26. 2022Self-Supervised Correspondence Estimation via Multiview Registrationcited by 5

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Register Any Point: Scaling 3D Point Cloud Registration by Flow Matching
  2. 2026PROSE: Training-Free Egocentric Scene Registration with Vision-Language Models
  3. 2025IDCNet: Guided Video Diffusion for Metric-Consistent RGBD Scene Generation with Precise Camera Control
  4. 2021Breaking Shortcut: Exploring Fully Convolutional Cycle-Consistency for Video Correspondence Learning
  5. 2026Learning Proposes, Geometry Disposes: A Modular Framework for Efficient Spatial Reasoning
  6. 2026Emergent Multi-View Geometry Through Self-Distillation
  7. 2026Demo-Pose: Depth-Monocular Modality Fusion For Object Pose Estimation
  8. 2026FS-I2P:A Hierarchical Focus-Sweep Registration Network with Dynamically Allocated Depth