Modelling Uncertainty in Deep Learning for Camera Relocalization

TL;DR

This paper introduces a real-time monocular camera relocalization system using Bayesian CNNs that estimates pose and uncertainty, improving accuracy and enabling scene presence detection in large-scale outdoor and indoor environments.

Contribution

It presents a novel Bayesian CNN approach for real-time 6-DOF relocalization that estimates uncertainty and enhances localization accuracy without extra engineering.

Findings

Achieves 2m and 6° accuracy outdoors, 0.5m and 10° indoors

Operates in under 6ms per inference, suitable for real-time applications

Uncertainty correlates with scene dissimilarity, aiding error estimation and scene detection

Abstract

We present a robust and real-time monocular six degree of freedom visual relocalization system. We use a Bayesian convolutional neural network to regress the 6-DOF camera pose from a single RGB image. It is trained in an end-to-end manner with no need of additional engineering or graph optimisation. The algorithm can operate indoors and outdoors in real time, taking under 6ms to compute. It obtains approximately 2m and 6 degrees accuracy for very large scale outdoor scenes and 0.5m and 10 degrees accuracy indoors. Using a Bayesian convolutional neural network implementation we obtain an estimate of the model's relocalization uncertainty and improve state of the art localization accuracy on a large scale outdoor dataset. We leverage the uncertainty measure to estimate metric relocalization error and to detect the presence or absence of the scene in the input image. We show that the model's uncertainty is caused by images being dissimilar to the training dataset in either pose or appearance.