mlsanityData to 14 Oct 2026

AffineGlue: Joint Matching and Robust Estimation

arXiv 2307.15381 ↗First posted 28 Jul 2023cs.CVcited by 5

Not on the map. Why?

Abstract (arXiv)

We propose AffineGlue, a method for joint two-view feature matching and robust estimation that reduces the combinatorial complexity of the problem by employing single-point minimal solvers. AffineGlue selects potential matches from one-to-many correspondences to estimate minimal models. Guided matching is then used to find matches consistent with the model, suffering less from the ambiguities of one-to-one matches. Moreover, we derive a new minimal solver for homography estimation, requiring only a single affine correspondence (AC) and a gravity prior. Furthermore, we train a neural network to reject ACs that are unlikely to lead to a good model. AffineGlue is superior to the SOTA on real-world datasets, even when assuming that the gravity direction points downwards. On PhotoTourism, the AUC@10deg score is improved by 6.6 points compared to the SOTA. On ScanNet, AffineGlue makes SuperPoint and SuperGlue achieve similar accuracy as the detector-free LoFTR.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Builds on 42 of its 101 references that are in the map. First posted on arXiv 28 Jul 2023.

Builds on

42 of 101 references.

References · 42

  1. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 7,005
  2. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,117
  3. 2021Vision Transformers for Dense Predictioncited by 1,803
  4. 2019Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfercited by 1,495
  5. 2019SuperGlue: Learning Feature Matching with Graph Neural Networkscited by 1,443
  6. 2017SuperPoint: Self-Supervised Interest Point Detection and Descriptioncited by 1,320
  7. 2015ORB-SLAM: a Versatile and Accurate Monocular SLAM Systemcited by 988
  8. 2021LoFTR: Detector-Free Local Feature Matching with Transformerscited by 984
  9. 2015NetVLAD: CNN architecture for weakly supervised place recognitioncited by 941
  10. 2020DISK: Learning local features with policy gradientcited by 312
  11. 2017HPatches: A benchmark and evaluation of handcrafted and learned local descriptorscited by 264
  12. 2016Large-Scale Image Retrieval with Attentive Deep Local Featurescited by 238
30 more
  1. 2020Image Matching across Wide Baselines: From Paper to Practicecited by 225
  2. 2017Working hard to know your neighbor's margins: Local descriptor learning losscited by 212
  3. 2019R2D2: Repeatable and Reliable Detector and Descriptorcited by 207
  4. 2017Learning a Multi-View Stereo Machinecited by 194
  5. 2017Learning to Find Good Correspondencescited by 176
  6. 2022ASpanFormer: Detector-Free Image Matching with Adaptive Span Transformercited by 171
  7. 2019MAGSAC++, a fast, reliable and accurate robust estimatorcited by 161
  8. 2019Learning Two-View Correspondences and Geometry Using Order-Aware Networkcited by 149
  9. 2019Point-Based Multi-View Stereo Networkcited by 147
  10. 2019SOSNet: Second Order Similarity Regularization for Local Descriptor Learningcited by 131
  11. 2019Neural-Guided RANSAC: Learning Where to Sample Model Hypothesescited by 129
  12. 2022MatchFormer: Interleaving Attention in Transformers for Feature Matchingcited by 126
  13. 2018MAGSAC: marginalizing sample consensuscited by 123
  14. 2019Key.Net: Keypoint Detection by Handcrafted and Learned CNN Filterscited by 118
  15. 2022MeshLoc: Mesh-Based Visual Localizationcited by 62
  16. 2016CNN Image Retrieval Learns from BoW: Unsupervised Fine-Tuning with Hard Examplescited by 58
  17. 2017Repeatability Is Not Enough: Learning Affine Regions via Discriminabilitycited by 57
  18. 2021Progressive Correspondence Pruning by Consensus Learningcited by 46
  19. 2019Large-scale, real-time visual-inertial localization revisitedcited by 45
  20. 2017Toward Geometric Deep SLAMcited by 38
  21. 2017A clever elimination strategy for efficient minimal solverscited by 17
  22. 2015Learning to Assign Orientations to Feature Pointscited by 16
  23. 2020Efficient Initial Pose-graph Generation for Global SfMcited by 14
  24. 2020Making Affine Correspondences Work in Camera Geometry Computationcited by 13
  25. 2019ACNe: Attentive Context Normalization for Robust Permutation-Equivariant Learningcited by 9
  26. 2022Self-Supervised Learning of Image Scale and Orientationcited by 9
  27. 2022NeFSAC: Neurally Filtered Minimal Samplescited by 9
  28. 2019Relative planar motion for vehicle-mounted cameras from a single affine correspondencecited by 8
  29. 2020Relative Pose from Deep Learned Depth and a Single Affine Correspondencecited by 8
  30. 2021Space-Partitioning RANSACcited by 6

Built on by

Cited by year

Citing papers per year: 2024: 1; 2025: 4.

5 papers.

  1. 2025Pow3R: Empowering Unconstrained 3D Reconstruction with Camera and Scene Priorscited by 72
  2. 2025POMATO: Marrying Pointmap Matching with Temporal Motion for Dynamic 3D Reconstructioncited by 25
  3. 2025Reconstructing 4D Spatial Intelligence: A Surveycited by 14
  4. 2024Affine steerers for structured keypoint descriptioncited by 6
  5. 2025SaLon3R: Structure-aware Long-term Generalizable 3D Reconstruction from Unposed Imagescited by 4