mlsanityData to 7 Oct 2026

Leveraging Equivariant Features for Absolute Pose Regression

arXiv 2204.02163 ↗First posted 5 Apr 2022cs.CVcited by 17

Not on the map. Why?

Abstract (arXiv)

While end-to-end approaches have achieved state-of-the-art performance in many perception tasks, they are not yet able to compete with 3D geometry-based methods in pose estimation. Moreover, absolute pose regression has been shown to be more related to image retrieval. As a result, we hypothesize that the statistical features learned by classical Convolutional Neural Networks do not carry enough geometric information to reliably solve this inherently geometric task. In this paper, we demonstrate how a translation and rotation equivariant Convolutional Neural Network directly induces representations of camera motions into the feature space. We then show that this geometric property allows for implicitly augmenting the training data under a whole group of image plane-preserving transformations. Therefore, we argue that directly learning equivariant features is preferable than learning data-intensive intermediate representations. Comprehensive experimental validation demonstrates that our lightweight model outperforms existing ones on standard datasets.

Key facts. Cited by 17 papers in the map (data through 7 Oct 2026). Builds on 32 of its 65 references that are in the map. First posted on arXiv 5 Apr 2022.

Builds on

32 of 65 references.

References · 32

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  3. 2014Going Deeper with Convolutionscited by 4,231
  4. 2016Group Equivariant Convolutional Networkscited by 1,154
  5. 2021Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gaugescited by 1,131
  6. 2017PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenescited by 768
  7. 2018Spherical CNNscited by 452
  8. 2019General E(2)-Equivariant Steerable CNNscited by 441
  9. 2016Steerable CNNscited by 298
  10. 2017Geometric Loss Functions for Camera Pose Regression with Deep Learningcited by 224
  11. 2018A General Theory of Equivariant CNNs on Homogeneous Spacescited by 223
  12. 2019Pix2Pose: Pixel-Wise Coordinate Regression of Objects for 6D Pose Estimationcited by 218
20 more
  1. 2017T-LESS: An RGB-D Dataset for 6D Pose Estimation of Texture-less Objectscited by 210
  2. 2021ReDet: A Rotation-equivariant Detector for Aerial Object Detectioncited by 202
  3. 2019Understanding the Limitations of CNN-based Absolute Camera Pose Regressioncited by 177
  4. 2017Geometry-Aware Learning of Maps for Camera Localizationcited by 161
  5. 2021GDRNPP: A Geometry-guided and Fully Learning-based Object Pose Estimatorcited by 148
  6. 2020EPOS: Estimating 6D Pose of Objects with Symmetriescited by 147
  7. 2020On Translation Invariance in CNNs: Convolutional Layers can Exploit Absolute Spatial Locationcited by 141
  8. 2016Image-based localization using LSTMs for structured feature correlationcited by 136
  9. 2015Modelling Uncertainty in Deep Learning for Camera Relocalizationcited by 133
  10. 2021Learning Multi-Scene Absolute Pose Regression with Transformerscited by 108
  11. 2018Segmentation-driven 6D Object Pose Estimationcited by 104
  12. 2019AtLoc: Attention Guided Camera Localizationcited by 92
  13. 2021Equivariant Point Network for 3D Point Cloud Analysiscited by 81
  14. 2017Image-based Localization using Hourglass Networkscited by 63
  15. 2014Deep Roto-Translation Scattering for Object Classificationcited by 49
  16. 2019Survey on Deep Neural Networks in Speech and Vision Systemscited by 37
  17. 2021Position, Padding and Predictions: A Deeper Look at Position Information in CNNscited by 26
  18. 2020Rotation Equivariant Siamese Networks for Trackingcited by 15
  19. 2018Cross-Domain 3D Equivariant Image Embeddingscited by 12
  20. 2021Paying Attention to Activation Maps in Camera Pose Regressioncited by 10

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 1; 2023: 7; 2024: 3; 2025: 5; 2026: 1.

17 papers.

Sort:

  1. 2023Neural Refinement for Absolute Pose Regression with Feature Synthesiscited by 37
  2. 2023A Survey on Deep Learning-Based Monocular Spacecraft Pose Estimation: Current State, Limitations and Prospectscited by 30
  3. 2023Graph-CoVis: GNN-based Multi-view Panorama Global Pose Estimationcited by 8
  4. 2023Loss it right: Euclidean and Riemannian Metrics in Learning-based Visual Odometrycited by 6
  5. 2023Self-Supervised Learning for Place Representation Generalization across Appearance Changescited by 4
  6. 2025Monocular visual simultaneous localization and mapping: (r)evolution from geometry to deep learning-based pipelinescited by 4
  7. 2025FastPose-ViT: A Vision Transformer for Real-Time Spacecraft Pose Estimationcited by 4
  8. 2022Fast and Lightweight Scene Regressor for Camera Relocalizationcited by 3
  9. 2023Densely Connected G-invariant Deep Neural Networks with Signed Permutation Representationscited by 2
  10. 2024Conditional Variational Autoencoders for Probabilistic Pose Regressioncited by 2
  11. 2024NaVIP: An Image-Centric Indoor Navigation Solution for Visually Impaired Peoplecited by 2
  12. 2024Activating Self-Attention for Multi-Scene Absolute Pose Regressioncited by 2
5 more
  1. 2023Impact of Disentanglement on Pruning Neural Networkscited by 1
  2. 2025Robust Visual Localization via Semantic-Guided Multi-Scale Transformercited by 0
  3. 2025ThermalLoc: A Vision Transformer-Based Approach for Robust Thermal Camera Relocalization in Large-Scale Environmentscited by 0
  4. 2025Equivariant symmetry-aware head pose estimation for fetal MRIcited by 0
  5. 2026Cross-Modal RGB-D Fusion Transformer for 6D Pose Estimation of Non-Cooperative Spacecraft with Stereo-Derived Depthcited by 0