Waypoint Models for Instruction-guided Navigation in Continuous Environments

TL;DR

This paper introduces language-conditioned waypoint prediction models for instruction-guided navigation in continuous environments, demonstrating improved success rates and efficiency over prior methods in the VLN-CE setting.

Contribution

The paper develops and evaluates a spectrum of waypoint prediction models, establishing new state-of-the-art results in instruction-guided navigation in continuous spaces.

Findings

More expressive models produce faster, simpler trajectories.

Lower-level actions better approximate shortest paths.

Achieved 4% higher success rate on VLN-CE leaderboard.

Abstract

Little inquiry has explicitly addressed the role of action spaces in language-guided visual navigation -- either in terms of its effect on navigation success or the efficiency with which a robotic agent could execute the resulting trajectory. Building on the recently released VLN-CE setting for instruction following in continuous environments, we develop a class of language-conditioned waypoint prediction networks to examine this question. We vary the expressivity of these models to explore a spectrum between low-level actions and continuous waypoint prediction. We measure task performance and estimated execution time on a profiled LoCoBot robot. We find more expressive models result in simpler, faster to execute trajectories, but lower-level actions can achieve better navigation metrics by approximating shortest paths better. Further, our models outperform prior work in VLN-CE and set a new state-of-the-art on the public leaderboard -- increasing success rate by 4% with our best model on this challenging task.