Hyp2Nav: Hyperbolic Planning and Curiosity for Crowd Navigation

TL;DR

Hyp2Nav introduces a hyperbolic learning approach for crowd navigation that improves efficiency, interpretability, and success rates by leveraging hyperbolic geometry to encode hierarchical decision-making in social environments.

Contribution

This work presents Hyp2Nav, a novel hyperbolic policy and curiosity model that enhances crowd navigation with fewer parameters and improved interpretability compared to traditional methods.

Findings

Achieves best success rates and returns in simulations.

Uses up to 6 times fewer parameters than state-of-the-art models.

Enables low-resource, interpretable crowd navigation in 2D embedding spaces.

Abstract

Autonomous robots are increasingly becoming a strong fixture in social environments. Effective crowd navigation requires not only safe yet fast planning, but should also enable interpretability and computational efficiency for working in real-time on embedded devices. In this work, we advocate for hyperbolic learning to enable crowd navigation and we introduce Hyp2Nav. Different from conventional reinforcement learning-based crowd navigation methods, Hyp2Nav leverages the intrinsic properties of hyperbolic geometry to better encode the hierarchical nature of decision-making processes in navigation tasks. We propose a hyperbolic policy model and a hyperbolic curiosity module that results in effective social navigation, best success rates, and returns across multiple simulation settings, using up to 6 times fewer parameters than competitor state-of-the-art models. With our approach, it becomes even possible to obtain policies that work in 2-dimensional embedding spaces, opening up new possibilities for low-resource crowd navigation and model interpretability. Insightfully, the internal hyperbolic representation of Hyp2Nav correlates with how much attention the robot pays to the surrounding crowds, e.g. due to multiple people occluding its pathway or to a few of them showing colliding plans, rather than to its own planned route. The code is available at https://github.com/GDam90/hyp2nav.