Submodular Optimization for Keyframe Selection & Usage in SLAM

TL;DR

This paper introduces new submodular optimization strategies for selecting keyframes in SLAM, reducing storage and computation while maintaining localization accuracy, and proposes a map summarization method for environment capture under size constraints.

Contribution

It presents novel submodular-based keyframe selection and submap generation methods that enhance SLAM efficiency and map summarization under size constraints.

Findings

Reduces number of saved keyframes

Improves per scan computation time

Maintains localization performance

Abstract

Keyframes are LiDAR scans saved for future reference in Simultaneous Localization And Mapping (SLAM), but despite their central importance most algorithms leave choices of which scans to save and how to use them to wasteful heuristics. This work proposes two novel keyframe selection strategies for localization and map summarization, as well as a novel approach to submap generation which selects keyframes that best constrain localization. Our results show that online keyframe selection and submap generation reduce the number of saved keyframes and improve per scan computation time without compromising localization performance. We also present a map summarization feature for quickly capturing environments under strict map size constraints.