Enhancing exploration algorithms for navigation with visual SLAM
This work addresses the challenge of integrating exploration algorithms with vision-based SLAM for robotic navigation, but it appears incremental as it focuses on enhancements rather than a fundamental breakthrough.
The paper tackled the problem of improving exploration algorithms for autonomous navigation by adapting them for use with visual SLAM, and the result showed evaluation of these enhancements in a photo-realistic simulator using both ground-truth and neural network-reconstructed depth maps to estimate exploration coverage.
Exploration is an important step in autonomous navigation of robotic systems. In this paper we introduce a series of enhancements for exploration algorithms in order to use them with vision-based simultaneous localization and mapping (vSLAM) methods. We evaluate developed approaches in photo-realistic simulator in two modes: with ground-truth depths and neural network reconstructed depth maps as vSLAM input. We evaluate standard metrics in order to estimate exploration coverage.