On learning visual odometry errors

Andrea De Maio, Simon Lacroix

TL;DR: This paper discusses different methods of pairing VO with deep learning and proposes a simultaneous prediction of corrections and uncertainty.

Abstract:

This paper fosters the idea that deep learning methods can be sided to classical
visual odometry pipelines to improve their accuracy and to produce uncertainty
models to their estimations. We show that the biases inherent to the visual odom-
etry process can be faithfully learnt and compensated for, and that a learning ar-
chitecture associated to a probabilistic loss function can jointly estimate a full
covariance matrix of the residual errors, defining a heteroscedastic error model.
Experiments on autonomous driving image sequences and micro aerial vehicles
camera acquisitions assess the possibility to concurrently improve visual odome-
try and estimate an error associated to its outputs.

Code: https://www.dropbox.com/s/xfa5wgl9k479mvm/on_learning_vo_errors.tar.gz?dl=0

Keywords: visual odometry, deep learning in robotics, uncertainty estimation in computer vision