mlsanityData to 7 Oct 2026

Generalized Differentiable RANSAC

arXiv 2212.13185 ↗First posted 26 Dec 2022cs.CVcited by 30

Not on the map. Why?

Abstract (arXiv)

We propose nabla-RANSAC, a generalized differentiable RANSAC that allows learning the entire randomized robust estimation pipeline. The proposed approach enables the use of relaxation techniques for estimating the gradients in the sampling distribution, which are then propagated through a differentiable solver. The trainable quality function marginalizes over the scores from all the models estimated within nabla-RANSAC to guide the network learning accurate and useful inlier probabilities or to train feature detection and matching networks. Our method directly maximizes the probability of drawing a good hypothesis, allowing us to learn better sampling distributions. We test nabla-RANSAC on various real-world scenarios on fundamental and essential matrix estimation, and 3D point cloud registration, outdoors and indoors, with handcrafted and learning-based features. It is superior to the state-of-the-art in terms of accuracy while running at a similar speed to its less accurate alternatives. The code and trained models are available at https://github.com/weitong8591/differentiableransac.

Key facts. Cited by 30 papers in the map (data through 7 Oct 2026). Builds on 31 of its 81 references that are in the map. First posted on arXiv 26 Dec 2022.

Builds on

31 of 81 references.

References · 31

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 6,979
  3. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  4. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,088
  5. 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,708
  6. 2019SuperGlue: Learning Feature Matching with Graph Neural Networkscited by 1,437
  7. 2017SuperPoint: Self-Supervised Interest Point Detection and Descriptioncited by 1,313
  8. 2016The Concrete Distribution: A Continuous Relaxation of Discrete Random Variablescited by 1,135
  9. 2021LoFTR: Detector-Free Local Feature Matching with Transformerscited by 974
  10. 20163DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructionscited by 333
  11. 2020PREDATOR: Registration of 3D Point Clouds with Low Overlapcited by 287
  12. 2016DSAC - Differentiable RANSAC for Camera Localizationcited by 264
19 more
  1. 2019Kornia: an Open Source Differentiable Computer Vision Library for PyTorchcited by 231
  2. 2022Geometric Transformer for Fast and Robust Point Cloud Registrationcited by 200
  3. 2017Learning to Find Good Correspondencescited by 176
  4. 2019MAGSAC++, a fast, reliable and accurate robust estimatorcited by 160
  5. 2019Learning Two-View Correspondences and Geometry Using Order-Aware Networkcited by 148
  6. 2020Understanding Gradient Clipping in Private SGD: A Geometric Perspectivecited by 140
  7. 2017Graph-Cut RANSACcited by 133
  8. 2019Neural-Guided RANSAC: Learning Where to Sample Model Hypothesescited by 129
  9. 2018MAGSAC: marginalizing sample consensuscited by 122
  10. 2019Stochastic Optimization of Sorting Networks via Continuous Relaxationscited by 111
  11. 2021Progressive Correspondence Pruning by Consensus Learningcited by 46
  12. 2015MODS: Fast and Robust Method for Two-View Matchingcited by 26
  13. 2021VSAC: Efficient and Accurate Estimator for H and Fcited by 17
  14. 2019Low-variance Black-box Gradient Estimates for the Plackett-Luce Distributioncited by 11
  15. 2020Learning Surrogates via Deep Embeddingcited by 11
  16. 2019ACNe: Attentive Context Normalization for Robust Permutation-Equivariant Learningcited by 9
  17. 2022NeFSAC: Neurally Filtered Minimal Samplescited by 9
  18. 2021FEDS -- Filtered Edit Distance Surrogatecited by 5
  19. 2016Hypergraph Modelling for Geometric Model Fittingcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 3; 2024: 12; 2025: 11; 2026: 4.

30 papers.

Sort:

By area:

  1. 2023Visual Geometry Grounded Deep Structure From Motioncited by 225
  2. 2024MASt3R-SfM: a Fully-Integrated Solution for Unconstrained Structure-from-Motioncited by 161
  3. 2024PF3plat: Pose-Free Feed-Forward 3D Gaussian Splattingcited by 60
  4. 2024GeoCalib: Learning Single-image Calibration with Geometric Optimizationcited by 50
  5. 2023Unifying Correspondence, Pose and NeRF for Pose-Free Novel View Synthesis from Stereo Pairscited by 37
  6. 2024Matching 2D Images in 3D: Metric Relative Pose from Metric Correspondencescited by 25
  7. 2025MP-SfM: Monocular Surface Priors for Robust Structure-from-Motioncited by 25
  8. 2025Emergent Outlier View Rejection in Visual Geometry Grounded Transformerscited by 19
  9. 2025Relative Pose Estimation through Affine Corrections of Monocular Depth Priorscited by 18
  10. 2024Learning to Make Keypoints Sub-Pixel Accuratecited by 13
  11. 2024FAR: Flexible, Accurate and Robust 6DoF Relative Camera Pose Estimationcited by 12
  12. 2025DaD: Distilled Reinforcement Learning for Diverse Keypoint Detectioncited by 12
18 more
  1. 2025TurboReg: TurboClique for Robust and Efficient Point Cloud Registrationcited by 12
  2. 2024HandDGP: Camera-Space Hand Mesh Prediction with Differentiable Global Positioningcited by 7
  3. 2025Deep Learning Reforms Image Matching: A Survey and Outlookcited by 7
  4. 2025SupeRANSAC: One RANSAC to Rule Them Allcited by 6
  5. 2024Image Matching Filtering and Refinement by Planes and Beyondcited by 5
  6. 2024Unsupervised Point Cloud Registration with Self-Distillationcited by 4
  7. 2024MinBackProp -- Backpropagating through Minimal Solverscited by 1
  8. 2024Efficient and Distributed Large-Scale 3D Map Registration using Tomographic Featurescited by 1
  9. 2025Monte Carlo Diffusion for Generalizable Learning-Based RANSACcited by 1
  10. 2023KVN: Keypoints Voting Network with Differentiable RANSAC for Stereo Pose Estimationcited by 0
  11. 2024Global Point Cloud Registration Network for Large Transformationscited by 0
  12. 2025A Birotation Solution for Relative Pose Problemscited by 0
  13. 2025Event-driven Robust Fitting on Neuromorphic Hardwarecited by 0
  14. 2025End2Reg: Learning Task-Specific Segmentation for Markerless Registration in Spine Surgerycited by 0
  15. 2026Weaving Light and Time: Unified Harmonic-Geometric Representation Learning for Dense RGB-Event Parsingcited by 0
  16. 2026Cross-View Feature Matching: Survey, Benchmarking, and Foundation-Model Perspectivescited by 0
  17. 2026DiffSAC: Diffusion-guided Sampling for Consensus-based Robust Estimationcited by 0
  18. 2026BLASt3R: Bundle Adjustment of Any Image Set with Multi-View Matching and Monocular Priorscited by 0