Search papers, labs, and topics across Lattice.
This paper introduces FreqNav, a novel framework for object-oriented aerial vision-and-language navigation that adapts visual token allocation based on the evolving perceptual priorities during navigation stages. By employing a Frequency Token Router and a Phase-dependent Grounding Module, FreqNav effectively shifts focus from low-frequency spatial layouts to high-frequency target details, minimizing interference from irrelevant objects. Experimental results indicate that FreqNav not only surpasses existing methods in performance but also achieves approximately three times faster inference, highlighting its practical applicability in real-world aerial autonomy scenarios.
A lightweight framework that dynamically reallocates visual tokens can enhance aerial navigation performance while tripling inference speed.
Object-oriented aerial vision-and-language navigation (VLN) requires searching for a described target and landing on it precisely, under long-horizon and closed-loop control. Guided by a target-descriptive instruction during navigation, perceptual priorities dynamically evolve: early-stage exploration prioritizes low-frequency spatial layout, and then shifts to high-frequency target details. Existing VLN methods model the varying perceptual requirements across navigation stages with identical visual tokens, leading to interference from irrelevant objects and background clutter. To this end, we therefore formulate long-horizon aerial navigation as a frequencypreference shift from spatial structure to local detail and propose FreqNav, a lightweight frequency-routing adaptive perception framework. Under a fixed computational budget, FreqNav dynamically reallocates visual tokens across frequency components according to the current navigation stage. A Frequency Token Router selects stage-relevant visual representations from dual-view observations, while a Phase-dependent Grounding Module anchors visual evidence through explicit supervision. A Diffusion Transformer then predicts smooth trajectories for continuous control. Experiments show that FreqNav outperforms strong baselines while achieving approximately 3x faster inference. Real-world deployment further demonstrates its effectiveness, efficiency, and practical potential for long-horizon aerial autonomy.
