mlsanityData to 7 Oct 2026

Geometric Loss Functions for Camera Pose Regression with Deep Learning

arXiv 1704.00390 ↗First posted 2 Apr 2017cs.CVcited by 224

Not on the map. Why?

Abstract (arXiv)

Deep learning has shown to be effective for robust and real-time monocular image relocalisation. In particular, PoseNet is a deep convolutional neural network which learns to regress the 6-DOF camera pose from a single image. It learns to localize using high level features and is robust to difficult lighting, motion blur and unknown camera intrinsics, where point based SIFT registration fails. However, it was trained using a naive loss function, with hyper-parameters which require expensive tuning. In this paper, we give the problem a more fundamental theoretical treatment. We explore a number of novel loss functions for learning camera pose which are based on geometry and scene reprojection error. Additionally we show how to automatically learn an optimal weighting to simultaneously regress position and orientation. By leveraging geometry, we demonstrate that our technique significantly improves PoseNet's performance across datasets ranging from indoor rooms to a small city.

Key facts. Cited by 224 papers in the map (data through 7 Oct 2026). First posted on arXiv 2 Apr 2017.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 45; 2022: 52; 2023: 41; 2024: 43; 2025: 27; 2026: 16.

224 papers.

Sort:

By area:

  1. 2021Back to the Feature: Learning Robust Camera Localization from Pixels to Posecited by 199
  2. 2022CroCo: Self-Supervised Pre-training for 3D Vision Tasks by Cross-View Completioncited by 158
  3. 2021GNeRF: GAN-based Neural Radiance Field without Posed Cameracited by 153
  4. 2021LENS: Localization enhanced by NeRF synthesiscited by 116
  5. 2022Map-free Visual Relocalization: Metric Pose Relative to a Single Imagecited by 112
  6. 2021Learning Multi-Scene Absolute Pose Regression with Transformerscited by 108
  7. 2023Accelerated Coordinate Encoding: Learning to Relocalize in Minutes using RGB and Posescited by 107
  8. 2022LaMAR: Benchmarking Localization and Mapping for Augmented Realitycited by 98
  9. 2023Deep Learning for Event-based Vision: A Comprehensive Survey and Benchmarkscited by 90
  10. 2022DFNet: Enhance Absolute Pose Regression with Direct Feature Matchingcited by 82
  11. 2024Reloc3r: Large-Scale Training of Relative Camera Pose Regression for Generalizable, Fast, and Accurate Visual Localizationcited by 81
  12. 2022ViewFormer: NeRF-free Neural Rendering from Few Images Using Transformerscited by 70
38 more
  1. 2022Multi-Frame Self-Supervised Depth with Transformerscited by 64
  2. 2021On the Limits of Pseudo Ground Truth in Visual Camera Re-localisationcited by 63
  3. 2022MeshLoc: Mesh-Based Visual Localizationcited by 62
  4. 2021Direct-PoseNet: Absolute Pose Regression with Photometric Consistencycited by 61
  5. 2022Is Geometry Enough for Matching in Visual Localization?cited by 48
  6. 2024GLACE: Global Local Accelerated Coordinate Encodingcited by 48
  7. 2021SOLD2: Self-supervised Occlusion-aware Line Description and Detectioncited by 45
  8. 2023Boosting 3-DoF Ground-to-Satellite Camera Localization Accuracy via Geometry-Guided Cross-View Transformercited by 45
  9. 2024Map-Relative Pose Regression for Visual Re-Localizationcited by 42
  10. 2021Full Surround Monodepth from Multiple Camerascited by 41
  11. 20233D Line Mapping Revisitedcited by 39
  12. 2021Visual Camera Re-Localization Using Graph Neural Networks and Relative Pose Supervisioncited by 37
  13. 2023Neural Refinement for Absolute Pose Regression with Feature Synthesiscited by 37
  14. 2022Visual Localization via Few-Shot Scene Region Classificationcited by 36
  15. 2021CoordiNet: uncertainty-aware pose regressor for reliable vehicle localizationcited by 34
  16. 2022DPODv2: Dense Correspondence-Based 6 DoF Pose Estimationcited by 32
  17. 2024The NeRFect Match: Exploring NeRF Features for Visual Localizationcited by 32
  18. 2021Projective Manifold Gradient Layer for Deep Rotation Regressioncited by 31
  19. 2024GS-CPR: Efficient Camera Pose Refinement via 3D Gaussian Splattingcited by 31
  20. 2021Learning to Predict 3D Lane Shape and Camera Pose from a Single Image via Geometry Constraintscited by 30
  21. 2023A Survey on Deep Learning-Based Monocular Spacecraft Pose Estimation: Current State, Limitations and Prospectscited by 30
  22. 2021VS-Net: Voting with Segmentation for Visual Localizationcited by 29
  23. 2024SceneGraphLoc: Cross-Modal Coarse Visual Localization on 3D Scene Graphscited by 29
  24. 2021Learning Camera Localization via Dense Scene Matchingcited by 28
  25. 2021LaLaLoc: Latent Layout Localisation in Dynamic, Unvisited Environmentscited by 28
  26. 2022Camera Pose Auto-Encoders for Improving Pose Regressioncited by 28
  27. 2023360Loc: A Dataset and Benchmark for Omnidirectional Visual Localization with Cross-device Queriescited by 28
  28. 2021Generalized Contrastive Optimization of Siamese Networks for Place Recognitioncited by 27
  29. 2022DiffPoseNet: Direct Differentiable Camera Pose Estimationcited by 27
  30. 2022Investigating the Role of Image Retrieval for Visual Localization -- An exhaustive benchmarkcited by 27
  31. 2023HSCNet++: Hierarchical Scene Coordinate Classification and Regression for Visual Localization with Transformercited by 26
  32. 2024Matching 2D Images in 3D: Metric Relative Pose from Metric Correspondencescited by 25
  33. 2024DiffusionNOCS: Managing Symmetry and Uncertainty in Sim2Real Multi-Modal Category-level Pose Estimationcited by 24
  34. 2023PNeRFLoc: Visual Localization with Point-based Neural Radiance Fieldscited by 22
  35. 2024The Unreasonable Effectiveness of Pre-Trained Features for Camera Pose Refinementcited by 21
  36. 2022Bimodal Camera Pose Prediction for Endoscopycited by 20
  37. 2023Visual Localization using Imperfect 3D Models from the Internetcited by 20
  38. 2025E-RayZer: Self-supervised 3D Reconstruction as Spatial Visual Pre-trainingcited by 20

50 most cited of 224 All 224 (JSON)