DF-SLAM: A Deep-Learning Enhanced Visual SLAM System based on Deep Local Features

TL;DR

DF-SLAM enhances traditional visual SLAM with deep learning-based local features, improving efficiency, stability, and robustness in diverse and challenging environments while maintaining real-time performance.

Contribution

It introduces a deep learning-based local feature descriptor into SLAM, replacing hand-crafted features, and demonstrates real-time performance and improved robustness.

Findings

Outperforms traditional SLAM in various scenes.

Maintains real-time operation on GPU.

Shows robustness in challenging illumination conditions.

Abstract

As the foundation of driverless vehicle and intelligent robots, Simultaneous Localization and Mapping(SLAM) has attracted much attention these days. However, non-geometric modules of traditional SLAM algorithms are limited by data association tasks and have become a bottleneck preventing the development of SLAM. To deal with such problems, many researchers seek to Deep Learning for help. But most of these studies are limited to virtual datasets or specific environments, and even sacrifice efficiency for accuracy. Thus, they are not practical enough. We propose DF-SLAM system that uses deep local feature descriptors obtained by the neural network as a substitute for traditional hand-made features. Experimental results demonstrate its improvements in efficiency and stability. DF-SLAM outperforms popular traditional SLAM systems in various scenes, including challenging scenes with intense illumination changes. Its versatility and mobility fit well into the need for exploring new environments. Since we adopt a shallow network to extract local descriptors and remain others the same as original SLAM systems, our DF-SLAM can still run in real-time on GPU.