mlsanityData to 14 Oct 2026

The 8-Point Algorithm as an Inductive Bias for Relative Pose Prediction by ViTs

arXiv 2208.08988 ↗First posted 18 Aug 2022cs.CVcited by 38

Not on the map. Why?

Abstract (arXiv)

We present a simple baseline for directly estimating the relative pose (rotation and translation, including scale) between two images. Deep methods have recently shown strong progress but often require complex or multi-stage architectures. We show that a handful of modifications can be applied to a Vision Transformer (ViT) to bring its computations close to the Eight-Point Algorithm. This inductive bias enables a simple method to be competitive in multiple settings, often substantially improving over the state of the art with strong performance gains in limited data regimes.

Key facts. Cited by 38 papers in the map (data through 14 Oct 2026). Builds on 50 of its 82 references that are in the map. First posted on arXiv 18 Aug 2022.

Builds on

50 of 82 references.

References · 50

  1. 2017Attention Is All You Needcited by 60,361
  2. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  3. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  4. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  5. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  6. 2019ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Taskscited by 2,222
  7. 2019Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfercited by 1,495
  8. 2019SuperGlue: Learning Feature Matching with Graph Neural Networkscited by 1,443
  9. 2017Matterport3D: Learning from RGB-D Data in Indoor Environmentscited by 1,376
  10. 2017SuperPoint: Self-Supervised Interest Point Detection and Descriptioncited by 1,320
  11. 2020NeRF in the Wild: Neural Radiance Fields for Unconstrained Photo Collectionscited by 1,232
  12. 2018On the Continuity of Rotation Representations in Neural Networkscited by 1,214
38 more
  1. 2019Habitat: A Platform for Embodied AI Researchcited by 1,191
  2. 2021Transformers in Vision: A Surveycited by 1,013
  3. 2015ORB-SLAM: a Versatile and Accurate Monocular SLAM Systemcited by 988
  4. 2021LoFTR: Detector-Free Local Feature Matching with Transformerscited by 984
  5. 2017PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volumecited by 968
  6. 2021DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Camerascited by 687
  7. 2021BARF: Bundle-Adjusting Neural Radiance Fieldscited by 642
  8. 2020TartanAir: A Dataset to Push the Limits of Visual SLAMcited by 506
  9. 2020DISK: Learning local features with policy gradientcited by 312
  10. 2022How Do Vision Transformers Work?cited by 307
  11. 2018Bilinear Attention Networkscited by 280
  12. 2016DSAC - Differentiable RANSAC for Camera Localizationcited by 266
  13. 2020Consistent Video Depth Estimationcited by 220
  14. 2018Stereo Magnification: Learning View Synthesis using Multiplane Imagescited by 211
  15. 2017Geometry-Aware Learning of Maps for Camera Localizationcited by 161
  16. 2020TartanVO: A Generalizable Learning-based VOcited by 108
  17. 2018InteriorNet: Mega-scale Multi-sensor Photo-realistic Indoor Scenes Datasetcited by 87
  18. 2017Camera Relocalization by Computing Pairwise Relative Poses Using Convolutional Neural Networkcited by 79
  19. 2018VLocNet++: Deep Multitask Learning for Semantic Visual Localization and Odometrycited by 61
  20. 2019To Learn or Not to Learn: Visual Localization from Essential Matricescited by 61
  21. 2017End-to-end weakly-supervised semantic alignmentcited by 60
  22. 2021Wide-Baseline Relative Camera Pose Estimation with Directional Learningcited by 49
  23. 2017Relative Camera Pose Estimation Using Convolutional Neural Networkscited by 47
  24. 2021UnsupervisedR&R: Unsupervised Point Cloud Registration via Differentiable Renderingcited by 44
  25. 2019The StreetLearn Environment and Datasetcited by 42
  26. 2021Extreme Rotation Estimation using Dense Correlation Volumescited by 42
  27. 2021Tangent Space Backpropagation for 3D Transformation Groupscited by 40
  28. 2021Planar Surface Reconstruction from Sparse Viewscited by 40
  29. 2018RPNet: an End-to-End Network for Relative Camera Pose Estimationcited by 31
  30. 2021Input-level Inductive Biases for 3D Reconstructioncited by 31
  31. 2022PlaneFormers: From Sparse View Planes to 3D Reconstructioncited by 26
  32. 2018Beyond Gr"obner Bases: Basis Selection for Minimal Solverscited by 25
  33. 2019Local Supports Global: Deep Camera Relocalization with Sequence Enhancementcited by 25
  34. 2015WxBS: Wide Baseline Stereo Generalizationscited by 23
  35. 2018Extreme Relative Pose Estimation for RGB-D Scans via Scene Completioncited by 22
  36. 2019Extreme Relative Pose Network under Hybrid Representationscited by 19
  37. 2020Associative3D: Volumetric Reconstruction from Sparse Viewscited by 18
  38. 2018Deep Fundamental Matrix Estimation without Correspondencescited by 15

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 5; 2023: 12; 2024: 10; 2025: 10; 2026: 1.

38 papers.

Sort:

By area:

  1. 2022SparseFusion: Distilling View-conditioned Diffusion for 3D Reconstructioncited by 216
  2. 2025FLARE: Feed-forward Geometry, Appearance and Camera Estimation from Uncalibrated Sparse Viewscited by 180
  3. 2023PF-LRM: Pose-Free Large Reconstruction Model for Joint Pose and Shape Predictioncited by 165
  4. 2024Cameras as Rays: Pose Estimation via Ray Diffusioncited by 131
  5. 2024RGBD Objects in the Wild: Scaling Real-World 3D Object Learning from RGB-D Videoscited by 115
  6. 2023RelPose++: Recovering 6D Poses from Sparse-view Observationscited by 87
  7. 2025Pow3R: Empowering Unconstrained 3D Reconstruction with Camera and Scene Priorscited by 72
  8. 2023LEAP: Liberate Sparse-view 3D Modeling from Camera Posescited by 65
  9. 2024PF3plat: Pose-Free Feed-Forward 3D Gaussian Splattingcited by 62
  10. 2022Few-View Object Reconstruction with Unknown Categories and Camera Posescited by 53
  11. 2023360-Degree Panorama Generation from Few Unregistered NFoV Imagescited by 42
  12. 2024SelfSplat: Pose-Free and 3D Prior-Free Generalizable 3D Gaussian Splattingcited by 42
26 more
  1. 2023Unifying Correspondence, Pose and NeRF for Pose-Free Novel View Synthesis from Stereo Pairscited by 37
  2. 2023POPE: 6-DoF Promptable Pose Estimation of Any Object, in Any Scene, with One Referencecited by 30
  3. 2022Last-Mile Embodied Visual Navigationcited by 25
  4. 2025E-RayZer: Self-supervised 3D Reconstruction as Spatial Visual Pre-trainingcited by 22
  5. 2023XVO: Generalized Visual Odometry via Cross-Modal Self-Trainingcited by 18
  6. 2024FAR: Flexible, Accurate and Robust 6DoF Relative Camera Pose Estimationcited by 12
  7. 2023Two-View Geometry Scoring Without Correspondencescited by 11
  8. 2022Leveraging Image Matching Toward End-to-End Relative Camera Pose Regressioncited by 6
  9. 2023PlaneRecTR++: Unified Query Learning for Joint 3D Planar Reconstruction and Pose Estimationcited by 6
  10. 2024Extreme Rotation Estimation in the Wildcited by 6
  11. 2025Emergent Extreme-View Geometry in 3D Foundation Modelscited by 6
  12. 2024SRPose: Two-view Relative Pose Estimation with Sparse Keypointscited by 5
  13. 2024Extreme Two-View Geometry From Object Poses with Diffusion Modelscited by 4
  14. 2025ZeroVO: Visual Odometry with Minimal Assumptionscited by 4
  15. 2023Learned Two-Plane Perspective Prior based Image Resampling for Efficient Object Detectioncited by 3
  16. 2024RESfM: Robust Deep Equivariant Structure from Motioncited by 2
  17. 2026Lost in Space? Vision-Language Models Struggle with Relative Camera Pose Estimationcited by 2
  18. 2022A Lightweight Domain Adaptive Absolute Pose Regressor Using Barlow Twins Objectivecited by 1
  19. 2024Generalizable Single-view Object Pose Estimation by Two-side Generating and Matchingcited by 1
  20. 2025Kaleidoscopic Background Attack: Disrupting Pose Estimation with Multi-Fold Radial Symmetry Texturescited by 1
  21. 2023Estimating Extreme 3D Image Rotation with Transformer Cross-Attentioncited by 0
  22. 2023LoCUS: Learning Multiscale 3D-consistent Features from Posed Imagescited by 0
  23. 2025Correspondence-Free Pose Estimation with Patterns: A Unified Approach for Multi-Dimensional Visioncited by 0
  24. 2025GCRayDiffusion: Pose-Free Surface Reconstruction via Geometric Consistent Ray Diffusioncited by 0
  25. 2025T-Graph: Enhancing Sparse-view Camera Pose Estimation by Pairwise Translation Graphcited by 0
  26. 2025iGaussian: Real-Time Camera Pose Estimation via Feed-Forward 3D Gaussian Splatting Inversioncited by 0