FreeFly-Thinking : Aligning Chain-of-Thought Reasoning with Continuous UAV Navigation

TL;DR

This paper introduces FreeFly-thinking, a novel UAV vision-language navigation framework that explicitly reasons through natural language instructions using chain-of-thought, improving robustness and efficiency in outdoor urban environments.

Contribution

It presents a new end-to-end VLN framework for UAVs that incorporates explicit reasoning, a dedicated UAV navigation dataset, and a two-stage training strategy, addressing gaps in outdoor scene navigation.

Findings

Strong performance on unseen tests

Robustness in complex outdoor environments

Efficient navigation in urban scenes

Abstract

Vision-Language Navigation aims to enable agents to understand natural language instructions and carry out appropriate navigation actions in real-world environments. Most work focuses on indoor settings, with little research in complex outdoor scenes. Current UAV Vision-and-Language Navigation models typically act as black boxes without explicit reasoning. We introduce FreeFly-thinking, an end-to-end VLN framework that converts the UAV agent's egocentric images and language instructions into a series of actions, inspired by environment of urban architecture proposed by OpenFly. We first construct a UAV dataset for navigation task, and then performing natural language chain of thought. We adopt a two-stage training strategy: Supervised fine-tuning and Reinforcement fine-tuning. Experiments on unseen test demonstrate a strong performance, presenting robustness and efficiency in UAV navigation issue.