Efficient deep learning-based semantic mapping approach using monocular vision for resource-limited mobile robots | Publicación