A Look at Improving Robustness in Visual-inertial SLAM by Moment Matching

TL;DR

This paper explores the use of moment matching via unscented Kalman filtering to enhance robustness in visual-inertial SLAM, addressing limitations of the traditional EKF under noisy and faulty data conditions.

Contribution

It introduces a moment matching approach for visual-inertial SLAM, improving robustness over standard EKF methods in challenging noise and data association scenarios.

Findings

Achieves state-of-the-art results on EuRoC MAV benchmark

Demonstrates improved robustness in noisy and faulty measurement conditions

Highlights advantages of unscented Kalman filtering in visual-inertial SLAM

Abstract

The fusion of camera sensor and inertial data is a leading method for ego-motion tracking in autonomous and smart devices. State estimation techniques that rely on non-linear filtering are a strong paradigm for solving the associated information fusion task. The de facto inference method in this space is the celebrated extended Kalman filter (EKF), which relies on first-order linearizations of both the dynamical and measurement model. This paper takes a critical look at the practical implications and limitations posed by the EKF, especially under faulty visual feature associations and the presence of strong confounding noise. As an alternative, we revisit the assumed density formulation of Bayesian filtering and employ a moment matching (unscented Kalman filtering) approach to both visual-inertial odometry and visual SLAM. Our results highlight important aspects in robustness both in dynamics propagation and visual measurement updates, and we show state-of-the-art results on EuRoC MAV drone data benchmark.