This repository contains examples of sensor registration using different manifolds and Lie groups. For the RGB-D visual odometry case, i.e., R^3 and SE(3), see: Continuous Direct Sparse Visual Odometry from RGB-D Images.
Continuous sensor registration is a new mathematical framework that enables nonparametric joint semantic/appearance and geometric representation of continuous functions using data. The joint semantic and geometric embedding is modeled by representing the processes in a reproducing kernel Hilbert space. The framework allows the functions to be defined on arbitrary smooth manifolds where the action of a Lie group is used to align them. The continuous functions allow the registration to be independent of a specific signal resolution and the framework is fully analytical with a closed-form derivation of the Riemannian gradient and Hessian.
- William Clark, Maani Ghaffari, Anthony Bloch. "Nonparametric Continuous Sensor Registration." Journal of Machine Learning Research 22.271 (2021): 1-50. http://jmlr.org/papers/v22/20-1468.html
@article{JMLR:v22:20-1468,
author = {William Clark and Maani Ghaffari and Anthony Bloch},
title = {Nonparametric Continuous Sensor Registration},
journal = {Journal of Machine Learning Research},
year = {2021},
volume = {22},
number = {271},
pages = {1-50},
url = {http://jmlr.org/papers/v22/20-1468.html}
}