PREDATOR: Registration of 3D Point Clouds with Low Overlap

2020/11/25 by Shengyu Huang, Žan Gojčič, Huang, Shengyu +8 · 43 citations
Computer Science · Earth and Planetary Sciences · Engineering · #3D Shape Modeling and Analysis #3D Surveying and Cultural Heritage #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #FOS: Electrical engineering #Image and Video Processing (eess.IV) #Robotics and Sensor-Based Localization #cs.CV #eess.IV #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.2011.13005

CVPR 2021 (Oral) - Improved performance after fixing GNN bug

openalex publication_date 2020/11/25 · arxiv created 2021/08/06 · arxiv updated 2021/08/09 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We introduce PREDATOR, a model for pairwise point-cloud registration with deep attention to the overlap region. Different from previous work, our model is specifically designed to handle (also) point-cloud pairs with low overlap. Its key novelty is an overlap-attention block for early information exchange between the latent encodings of the two point clouds. In this way the subsequent decoding of the latent representations into per-point features is conditioned on the respective other point cloud, and thus can predict which points are not only salient, but also lie in the overlap region between the two point clouds. The ability to focus on points that are relevant for matching greatly improves performance: PREDATOR raises the rate of successful registrations by more than 20% in the low-overlap scenario, and also sets a new state of the art for the 3DMatch benchmark with 89% registration recall.

Citations

• A Relationship Between Arbitrary Positive Matrices and Doubly Stochastic Matrices
• Instance Normalization: The Missing Ingredient for Fast Stylization
• Semi-Supervised Classification with Graph Convolutional Networks
• Neural Message Passing for Quantum Chemistry
• R2D2: Repeatable and Reliable Detector and Descriptor

Cited by

• MCI-Net: A Robust Multi-Domain Context Integration Network for Point Cloud Registration
• T2LDM++: A Self-Conditioned Representation Guided Diffusion Model for Realistic Text-to-LiDAR Scene Generation
• End2Reg: Learning Task-Specific Segmentation for Markerless Registration in Spine Surgery
• Geometry-to-Image Synthesis-Driven Generative Point Cloud Registration
• FUSER: Feed-Forward MUltiview 3D Registration Transformer and SE(3)^N Diffusion Refinement
• Attention-guided reference point shifting for Gaussian-mixture-based partial point set registration
• Register Any Point: Scaling 3D Point Cloud Registration by Flow Matching
• LAHNet: Local Attentive Hashing Network for Point Cloud Registration
• PointCNN++: Performant Convolution on Native Points
• ViGG: Robust RGB-D Point Cloud Registration using Visual-Geometric Mutual Guidance
• The Joint Gromov Wasserstein Objective for Multiple Object Matching
• CoordAR: One-Reference 6D Pose Estimation of Novel Objects via Autoregressive Coordinate Map Generation
• NeuralBoneReg: An Instance-Specific Label-Free Point Cloud-Based Method for Multi-Modal Bone Surface Registration
• WetExplorer: Automating Wetland Greenhouse-Gas Surveys with an Autonomous Mobile Robot
• Adaptive Agent Selection and Interaction Network for Image-to-point cloud Registration
• Monocular Visual 8D Pose Estimation for Articulated Bicycles and Cyclists
• Leveraging Cycle-Consistent Anchor Points for Self-Supervised RGB-D Registration
• DINOReg: Strong Point Cloud Registration with Vision Foundation Model
• SGAligner++: Cross-Modal Language-Aided 3D Scene Graph Alignment
• Robust Point Cloud Registration via Geometric Overlapping Guided Rotation Search
• DualReg: Dual-Space Filtering and Reinforcement for Rigid Registration
• iMatcher: Improve matching in point cloud registration via local-to-global geometric consistency learning
• Cross3DReg: Towards a Large-scale Real-world Cross-source Point Cloud Registration Benchmark
• Delving into Dynamic Scene Cue-Consistency for Robust 3D Multi-Object Tracking
• Combinative Matching for Geometric Shape Assembly
• Correspondence-Free Fast and Robust Spherical Point Pattern Registration
• Modality-Aware Feature Matching: A Comprehensive Review of Single- and Cross-Modality Techniques
• PIVOTS: Aligning unseen Structures using Preoperative to Intraoperative Volume-To-Surface Registration for Liver Navigation
• RARE: Refine Any Registration of Pairwise Point Clouds via Zero-Shot Learning
• Diff²I2P: Differentiable Image-to-Point Cloud Registration with Diffusion Prior
• TurboReg: TurboClique for Robust and Efficient Point Cloud Registration
• Single-Frame Point-Pixel Registration via Supervised Cross-Modal Feature Matching
• CA-I2P: Channel-Adaptive Registration Network with Global Optimal Selection
• Dense 3D Displacement Estimation for Landslide Monitoring via Fusion of TLS Point Clouds and Embedded RGB Images
• MT-PCR: Hybrid Mamba-Transformer Network with Spatial Serialization for Point Cloud Registration
• Linearly Solving Robust Rotation Estimation
• Rectified Point Flow: Generic Point Cloud Pose Estimation
• SupeRANSAC: One RANSAC to Rule Them All
• Equivariant Flow Matching for Point Cloud Assembly
• GMatch: A Lightweight, Geometry-Constrained Keypoint Matcher for Zero-Shot 6DoF Pose Estimation in Robotic Grasp Tasks
• Cross-modal feature fusion for robust point cloud registration with ambiguous geometry
• 3D Hand-Eye Calibration for Collaborative Robot Arm: Look at Robot Base Once
• Landmark-Free Preoperative-to-Intraoperative Registration in Laparoscopic Liver Resection

Related

• Fast and Robust Registration of Partially Overlapping Point Clouds
• RoCNet: 3D Robust Registration of Point-Clouds using Deep Learning
• Point Cloud Registration using Representative Overlapping Points
• Deep Models with Fusion Strategies for MVP Point Cloud Registration
• OAAFormer: Robust and Efficient Point Cloud Registration Through Overlapping-Aware Attention in Transformer