# 3D Local Features for Direct Pairwise Registration

arXiv: https://arxiv.org/abs/1904.04281 · HTML: https://mlsanity.com/paper/1904.04281/

## Key facts

- Cited by 26 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 8 Apr 2019.

## Built on by (most cited first, 26 of 26)

- [PointDSC: Robust Point Cloud Registration using Deep Spatial Consistency](https://mlsanity.com/paper/2103.05465/) (2021)
- [Density-aware Chamfer Distance as a Comprehensive Metric for Point Cloud Completion](https://mlsanity.com/paper/2111.12702/) (2021)
- [A comprehensive survey on point cloud registration](https://mlsanity.com/paper/2103.02690/) (2021)
- [You Only Hypothesize Once: Point Cloud Registration with Rotation-equivariant Descriptors](https://mlsanity.com/paper/2109.00182/) (2021)
- [Point-set Distances for Learning Representations of 3D Point Clouds](https://mlsanity.com/paper/2102.04014/) (2021)
- [UnsupervisedR&R: Unsupervised Point Cloud Registration via Differentiable Rendering](https://mlsanity.com/paper/2102.11870/) (2021)
- [Learning general and distinctive 3D local deep descriptors for point cloud registration](https://mlsanity.com/paper/2105.10382/) (2021)
- [Multiway Non-rigid Point Cloud Registration via Learned Functional Map Synchronization](https://mlsanity.com/paper/2111.12878/) (2021)
- [Bootstrap Your Own Correspondences](https://mlsanity.com/paper/2106.00677/) (2021)
- [Self-Supervised Point Cloud Registration with Deep Versatile Descriptors](https://mlsanity.com/paper/2201.10034/) (2022)
- [Register Any Point: Scaling 3D Point Cloud Registration by Flow Matching](https://mlsanity.com/paper/2512.01850/) (2025)
- [RPSRNet: End-to-End Trainable Rigid Point Set Registration Network using Barnes-Hut 2D-Tree Representation](https://mlsanity.com/paper/2104.05328/) (2021)
- [VRNet: Learning the Rectified Virtual Corresponding Points for 3D Point Cloud Registration](https://mlsanity.com/paper/2203.13241/) (2022)
- [WSDesc: Weakly Supervised 3D Local Descriptor Learning for Point Cloud Registration](https://mlsanity.com/paper/2108.02740/) (2021)
- [Self-Supervised Correspondence Estimation via Multiview Registration](https://mlsanity.com/paper/2212.03236/) (2022)
- [Loss Distillation via Gradient Matching for Point Cloud Completion with Weighted Chamfer Distance](https://mlsanity.com/paper/2409.06171/) (2024)
- [3D3L: Deep Learned 3D Keypoint Detection and Description for LiDARs](https://mlsanity.com/paper/2103.13808/) (2021)
- [A Representation Separation Perspective to Correspondences-free Unsupervised 3D Point Cloud Registration](https://mlsanity.com/paper/2203.13239/) (2022)
- [Accurate Instance-Level CAD Model Retrieval in a Large-Scale Database](https://mlsanity.com/paper/2207.01339/) (2022)
- [Learning a Task-specific Descriptor for Robust Matching of 3D Point Clouds](https://mlsanity.com/paper/2210.14899/) (2022)
- [(Just) A Spoonful of Refinements Helps the Registration Error Go Down](https://mlsanity.com/paper/2108.03257/) (2021)
- [Overlap Bias Matching is Necessary for Point Cloud Registration](https://mlsanity.com/paper/2308.09364/) (2023)
- [Partial Transport for Point-Cloud Registration](https://mlsanity.com/paper/2309.15787/) (2023)
- [3D Registration in 30 Years: A Survey](https://mlsanity.com/paper/2412.13735/) (2024)
- [Accurate Ground-Truth Depth Image Generation via Overfit Training of Point Cloud Registration using Local Frame Sets](https://mlsanity.com/paper/2207.07016/) (2022)
- [Correspondence-Free Pose Estimation with Patterns: A Unified Approach for Multi-Dimensional Vision](https://mlsanity.com/paper/2503.00051/) (2025)

## Abstract (arXiv)

We present a novel, data driven approach for solving the problem of registration of two point cloud scans. Our approach is direct in the sense that a single pair of corresponding local patches already provides the necessary transformation cue for the global registration. To achieve that, we first endow the state of the art PPF-FoldNet auto-encoder (AE) with a pose-variant sibling, where the discrepancy between the two leads to pose-specific descriptors. Based upon this, we introduce RelativeNet, a relative pose estimation network to assign correspondence-specific orientations to the keypoints, eliminating any local reference frame computations. Finally, we devise a simple yet effective hypothesize-and-verify algorithm to quickly use the predictions and align two point sets. Our extensive quantitative and qualitative experiments suggests that our approach outperforms the state of the art in challenging real datasets of pairwise registration and that augmenting the keypoints with local pose information leads to better generalization and a dramatic speed-up.