Guardado en:
Detalles Bibliográficos
Autores principales: Zangeneh, Fereidoon, Dekel, Amit, Pieropan, Alessandro, Jensfelt, Patric
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2504.07260
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866915235686776832
author Zangeneh, Fereidoon
Dekel, Amit
Pieropan, Alessandro
Jensfelt, Patric
author_facet Zangeneh, Fereidoon
Dekel, Amit
Pieropan, Alessandro
Jensfelt, Patric
contents Visual relocalization is the task of estimating the camera pose given an image it views. Absolute pose regression offers a solution to this task by training a neural network, directly regressing the camera pose from image features. While an attractive solution in terms of memory and compute efficiency, absolute pose regression's predictions are inaccurate and unreliable outside the training domain. In this work, we propose a novel method for quantifying the epistemic uncertainty of an absolute pose regression model by estimating the likelihood of observations within a variational framework. Beyond providing a measure of confidence in predictions, our approach offers a unified model that also handles observation ambiguities, probabilistically localizing the camera in the presence of repetitive structures. Our method outperforms existing approaches in capturing the relation between uncertainty and prediction error.
format Preprint
id arxiv_https___arxiv_org_abs_2504_07260
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Quantifying Epistemic Uncertainty in Absolute Pose Regression
Zangeneh, Fereidoon
Dekel, Amit
Pieropan, Alessandro
Jensfelt, Patric
Computer Vision and Pattern Recognition
Visual relocalization is the task of estimating the camera pose given an image it views. Absolute pose regression offers a solution to this task by training a neural network, directly regressing the camera pose from image features. While an attractive solution in terms of memory and compute efficiency, absolute pose regression's predictions are inaccurate and unreliable outside the training domain. In this work, we propose a novel method for quantifying the epistemic uncertainty of an absolute pose regression model by estimating the likelihood of observations within a variational framework. Beyond providing a measure of confidence in predictions, our approach offers a unified model that also handles observation ambiguities, probabilistically localizing the camera in the presence of repetitive structures. Our method outperforms existing approaches in capturing the relation between uncertainty and prediction error.
title Quantifying Epistemic Uncertainty in Absolute Pose Regression
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2504.07260