mlsanityData to 7 Oct 2026

A Probabilistic Framework for Visual Localization in Ambiguous Scenes

arXiv 2301.02086 ↗First posted 5 Jan 2023cs.CV cs.ROcited by 12

Not on the map. Why?

Abstract (arXiv)

Visual localization allows autonomous robots to relocalize when losing track of their pose by matching their current observation with past ones. However, ambiguous scenes pose a challenge for such systems, as repetitive structures can be viewed from many distinct, equally likely camera poses, which means it is not sufficient to produce a single best pose hypothesis. In this work, we propose a probabilistic framework that for a given image predicts the arbitrarily shaped posterior distribution of its camera pose. We do this via a novel formulation of camera pose regression using variational inference, which allows sampling from the predicted distribution. Our method outperforms existing methods on localization in ambiguous scenes. Code and data will be released at https://github.com/efreidun/vapor.

Key facts. Cited by 12 papers in the map (data through 7 Oct 2026). Builds on 24 of its 39 references that are in the map. First posted on arXiv 5 Jan 2023.

Builds on

24 of 39 references.

References · 24

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  4. 2013Auto-Encoding Variational Bayescited by 3,980
  5. 2018On the Continuity of Rotation Representations in Neural Networkscited by 1,186
  6. 2015NetVLAD: CNN architecture for weakly supervised place recognitioncited by 935
  7. 2021Neural Fields in Visual Computing and Beyondcited by 539
  8. 2021Patch-NetVLAD: Multi-Scale Fusion of Locally-Global Descriptors for Place Recognitioncited by 236
  9. 2017Geometric Loss Functions for Camera Pose Regression with Deep Learningcited by 224
  10. 2021Back to the Feature: Learning Robust Camera Localization from Pixels to Posecited by 199
  11. 2019Understanding the Limitations of CNN-based Absolute Camera Pose Regressioncited by 177
  12. 2017Geometry-Aware Learning of Maps for Camera Localizationcited by 161
12 more
  1. 2016Image-based localization using LSTMs for structured feature correlationcited by 136
  2. 2015Modelling Uncertainty in Deep Learning for Camera Relocalizationcited by 133
  3. 2021LENS: Localization enhanced by NeRF synthesiscited by 116
  4. 2017Deep Learning Features at Scale for Visual Place Recognitioncited by 96
  5. 2019Overcoming Limitations of Mixture Density Networks: A Sampling and Fitting Framework for Multimodal Future Predictioncited by 96
  6. 2021Implicit-PDF: Non-Parametric Representation of Probability Distributions on the Rotation Manifoldcited by 65
  7. 2021Direct-PoseNet: Absolute Pose Regression with Photometric Consistencycited by 61
  8. 2017VidLoc: A Deep Spatio-Temporal Model for 6-DoF Video-Clip Relocalizationcited by 58
  9. 2020Deep Bingham Networks: Dealing with Uncertainty and Ambiguity in Pose Estimationcited by 43
  10. 2021CoordiNet: uncertainty-aware pose regressor for reliable vehicle localizationcited by 34
  11. 2021Reassessing the Limitations of CNN Methods for Camera Pose Regressioncited by 19
  12. 2021MILIOM: Tightly Coupled Multi-Input Lidar-Inertia Odometry and Mappingcited by 1

Built on by

Cited by year (2026 to 7 Oct)

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

12 papers.

  1. 2024Around the World in 80 Timesteps: A Generative Approach to Global Visual Geolocationcited by 23
  2. 2024HR-APR: APR-agnostic Framework with Uncertainty Estimation and Hierarchical Refinement for Camera Relocalisationcited by 16
  3. 2023Fusing Structure from Motion and Simulation-Augmented Pose Regression from Optical Flow for Challenging Indoor Environmentscited by 5
  4. 2026EAGOR: Embodied Reasoning in Omni-directioncited by 3
  5. 2024Conditional Variational Autoencoders for Probabilistic Pose Regressioncited by 2
  6. 2024MobileARLoc: On-device Robust Absolute Localisation for Pervasive Markerless Mobile ARcited by 1
  7. 2023Robust Localization with Visual-Inertial Odometry Constraints for Markerless Mobile ARcited by 0
  8. 2023KS-APR: Keyframe Selection for Robust Absolute Pose Regressioncited by 0
  9. 2024PoseINN: Realtime Visual-based Pose Regression and Localization with Invertible Neural Networkscited by 0
  10. 2025Quantifying Epistemic Uncertainty in Absolute Pose Regressioncited by 0
  11. 2025Semantic and Feature Guided Uncertainty Quantification of Visual Localization for Autonomous Vehiclescited by 0
  12. 2025Visible Structure Retrieval for Lightweight Image-Based Relocalisationcited by 0