mlsanityData to 7 Oct 2026

ObjectMatch: Robust Registration using Canonical Object Correspondences

arXiv 2212.01985 ↗First posted 5 Dec 2022cs.CV cs.LGcited by 15

Not on the map. Why?

Abstract (arXiv)

We present ObjectMatch, a semantic and object-centric camera pose estimator for RGB-D SLAM pipelines. Modern camera pose estimators rely on direct correspondences of overlapping regions between frames; however, they cannot align camera frames with little or no overlap. In this work, we propose to leverage indirect correspondences obtained via semantic object identification. For instance, when an object is seen from the front in one frame and from the back in another frame, we can provide additional pose constraints through canonical object correspondences. We first propose a neural network to predict such correspondences on a per-pixel level, which we then combine in our energy formulation with state-of-the-art keypoint matching solved with a joint Gauss-Newton optimization. In a pairwise setting, our method improves registration recall of state-of-the-art feature matching, including from 24% to 45% in pairs with 10% or less inter-frame overlap. In registering RGB-D sequences, our method outperforms cutting-edge SLAM baselines in challenging, low-frame-rate scenarios, achieving more than 35% reduction in trajectory error in multiple scenes.

Key facts. Cited by 15 papers in the map (data through 7 Oct 2026). Builds on 36 of its 51 references that are in the map. First posted on arXiv 5 Dec 2022.

Builds on

36 of 51 references.

References · 36

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  3. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  4. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,441
  5. 2017Mask R-CNNcited by 7,259
  6. 2016Feature Pyramid Networks for Object Detectioncited by 5,021
  7. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,088
  8. 2019SuperGlue: Learning Feature Matching with Graph Neural Networkscited by 1,437
  9. 2017SuperPoint: Self-Supervised Interest Point Detection and Descriptioncited by 1,313
  10. 2016Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Networkcited by 1,199
  11. 2021LoFTR: Detector-Free Local Feature Matching with Transformerscited by 974
  12. 2016ORB-SLAM2: an Open-Source SLAM System for Monocular, Stereo and RGB-D Camerascited by 801
24 more
  1. 2020Accelerating 3D Deep Learning with PyTorch3Dcited by 781
  2. 2018Open3D: A Modern Library for 3D Data Processingcited by 717
  3. 2021NICE-SLAM: Neural Implicit Scalable Encoding for SLAMcited by 586
  4. 2019Normalized Object Coordinate Space for Category-Level 6D Object Pose and Size Estimationcited by 490
  5. 2021iMAP: Implicit Mapping and Positioning in Real-Timecited by 476
  6. 2014Deep metric learning using Triplet networkcited by 421
  7. 20163DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructionscited by 333
  8. 2020PREDATOR: Registration of 3D Point Clouds with Low Overlapcited by 287
  9. 2020Deep Global Registrationcited by 228
  10. 2022Geometric Transformer for Fast and Robust Point Cloud Registrationcited by 200
  11. 2018Scan2CAD: Learning CAD Model Alignment in RGB-D Scanscited by 142
  12. 2016BundleFusion: Real-time Globally Consistent 3D Reconstruction using On-the-fly Surface Re-integrationcited by 116
  13. 2020DI-Fusion: Online Implicit 3D Reconstruction with Deep Priorscited by 70
  14. 2021ROCA: Robust CAD Model Retrieval and Alignment from a Single Imagecited by 66
  15. 2020NodeSLAM: Neural Object Descriptors for Multi-View Shape Reconstructioncited by 61
  16. 2015Multimodal Deep Learning for Robust RGB-D Object Recognitioncited by 56
  17. 2019End-to-End CAD Model Retrieval and 9DoF Alignment in 3D Scanscited by 40
  18. 2020Vid2CAD: CAD Model Alignment using Multi-View Constraints from Videoscited by 33
  19. 2022Virtual Correspondence: Humans as a Cue for Extreme-View Geometrycited by 25
  20. 2020Seeing Behind Objects for 3D Multi-Object Tracking in RGB-D Sequencescited by 13
  21. 2018PlaneMatch: Patch Coplanarity Prediction for Robust RGB-D Reconstructioncited by 8
  22. 2021Object-Augmented RGB-D SLAM for Wide-Disparity Relocalisationcited by 7
  23. 2021Visual Correspondence Hallucinationcited by 5
  24. 2019re-OBJ: Jointly Learning the Foreground and Background for Object Instance Re-identificationcited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 3; 2024: 2; 2025: 5; 2026: 5.

15 papers.

Sort:

By area:

  1. 2025Any6D: Model-free 6D Pose Estimation of Novel Objectscited by 47
  2. 2023Open-vocabulary object 6D pose estimationcited by 29
  3. 2025One2Any: One-Reference 6D Pose Estimation for Any Objectcited by 17
  4. 2024RTG-SLAM: Real-time 3D Reconstruction at Scale using Gaussian Splattingcited by 10
  5. 2024High-resolution open-vocabulary object 6D pose estimationcited by 10
  6. 2023LFM-3D: Learnable Feature Matching Across Wide Baselines Using 3D Signalscited by 8
  7. 2023High-Throughput and Accurate 3D Scanning of Cattle Using Time-of-Flight Sensors and Deep Learningcited by 2
  8. 2025Co-VisiON: Co-Visibility ReasONing on Sparse Image Sets of Indoor Scenescited by 2
  9. 2025CoordAR: One-Reference 6D Pose Estimation of Novel Objects via Autoregressive Coordinate Map Generationcited by 1
  10. 2025FOCUS -- Multi-View Foot Reconstruction From Synthetically Trained Dense Correspondencescited by 0
  11. 2026Learning Fine-Grained Correspondence with Cross-Perspective Perception for Open-Vocabulary 6D Object Pose Estimationcited by 0
  12. 2026Object Pose Transformer: Unifying Unseen Object Pose Estimationcited by 0
3 more
  1. 2026Semantic Prior Guided One-View 6D Pose Estimation for Novel Objectscited by 0
  2. 2026Learning Cross-View Semantic Priors for Single-Reference Unseen Object Pose Estimationcited by 0
  3. 2026Pose Anything Anywhere:Model-free Object Poses from Arbitrary Referencescited by 0