Learning to SLAM on the Fly in Unknown Environments: A Continual Learning Approach for Drones in Visually Ambiguous Scenes

TL;DR

This paper introduces a novel continual learning approach for drone SLAM in unknown, visually ambiguous environments, using a modulated dictionary learning pipeline with Bayesian surprise to enable on-the-fly adaptation.

Contribution

It proposes a new method combining dictionary learning and Bayesian surprise for real-time SLAM adaptation in unseen environments, addressing limitations of offline-trained neural networks.

Findings

Effective in challenging warehouse scenarios with ambiguous scenes

Enables on-the-fly adaptation for drones in unknown environments

Improves visual disambiguation in SLAM systems

Abstract

Learning to safely navigate in unknown environments is an important task for autonomous drones used in surveillance and rescue operations. In recent years, a number of learning-based Simultaneous Localisation and Mapping (SLAM) systems relying on deep neural networks (DNNs) have been proposed for applications where conventional feature descriptors do not perform well. However, such learning-based SLAM systems rely on DNN feature encoders trained offline in typical deep learning settings. This makes them less suited for drones deployed in environments unseen during training, where continual adaptation is paramount. In this paper, we present a new method for learning to SLAM on the fly in unknown environments, by modulating a low-complexity Dictionary Learning and Sparse Coding (DLSC) pipeline with a newly proposed Quadratic Bayesian Surprise (QBS) factor. We experimentally validate our approach with data collected by a drone in a challenging warehouse scenario, where the high number of ambiguous scenes makes visual disambiguation hard.