Learning Social Navigation from Demonstrations with Deep Neural Networks

TL;DR

This paper presents a deep learning framework for social navigation, enabling robots to plan paths that reach targets while avoiding obstacles and respecting human environments, using separate models for global and local planning.

Contribution

The study introduces a dual deep learning model approach for human-aware navigation, improving path planning performance over traditional neural network methods.

Findings

Successfully achieved human-aware navigation in simulation

Generated paths that reach targets while avoiding obstacles

Outperformed feed-forward neural networks in path planning

Abstract

Traditional path-planning techniques treat humans as obstacles. This has changed since robots started to enter human environments. On modern robots, social navigation has become an important aspect of navigation systems. To use learning-based techniques to achieve social navigation, a powerful framework that is capable of representing complex functions with as few data as possible is required. In this study, we benefited from recent advances in deep learning at both global and local planning levels to achieve human-aware navigation on a simulated robot. Two distinct deep models are trained with respective objectives: one for global planning and one for local planning. These models are then employed in the simulated robot. In the end, it has been shown that our model can successfully carry out both global and local planning tasks. We have shown that our system could generate paths that successfully reach targets while avoiding obstacles with better performance compared to feed-forward neural networks.