Hyperion -- A fast, versatile symbolic Gaussian Belief Propagation framework for Continuous-Time SLAM

TL;DR

Hyperion introduces a fast, versatile Gaussian Belief Propagation framework for continuous-time SLAM, enabling decentralized probabilistic inference and significantly improving computational efficiency over previous methods.

Contribution

The paper presents Hyperion, a novel GBP framework for CTSLAM that achieves high speedups and supports decentralized inference, addressing computational challenges in multi-sensor SLAM.

Findings

Achieved speedups between 2.43x and 110.31x over previous implementations.

Demonstrated effective motion tracking and localization in empirical tests.

Validated the framework's scalability and robustness through ablation studies.

Abstract

Continuous-Time Simultaneous Localization And Mapping (CTSLAM) has become a promising approach for fusing asynchronous and multi-modal sensor suites. Unlike discrete-time SLAM, which estimates poses discretely, CTSLAM uses continuous-time motion parametrizations, facilitating the integration of a variety of sensors such as rolling-shutter cameras, event cameras and Inertial Measurement Units (IMUs). However, CTSLAM approaches remain computationally demanding and are conventionally posed as centralized Non-Linear Least Squares (NLLS) optimizations. Targeting these limitations, we not only present the fastest SymForce-based [Martiros et al., RSS 2022] B- and Z-Spline implementations achieving speedups between 2.43x and 110.31x over Sommer et al. [CVPR 2020] but also implement a novel continuous-time Gaussian Belief Propagation (GBP) framework, coined Hyperion, which targets decentralized probabilistic inference across agents. We demonstrate the efficacy of our method in motion tracking and localization settings, complemented by empirical ablation studies.