Parse, Search, and Confirmation: Training-Free Aerial Vision-and-Dialog Navigation with Chain-of-Thought Reasoning and Structured Spatial Memory

2026-07-13

AI Summary

This work addresses the unreliability of vision-and-dialogue navigation for high-altitude unmanned aerial vehicles under zero-shot conditions, primarily caused by weak directional understanding and the absence of explicit spatial memory. To overcome these limitations, the authors propose the PSC-AVDN framework, which introduces—within a zero-shot setting—a structured geometric and visual spatial memory through a three-stage Parse–Search–Confirm reasoning pipeline. The framework incorporates a dual-chain-of-thought mechanism: a Search Chain (S-CoT) for geometrically parsing ambiguous linguistic instructions and a Confirmation Chain (C-CoT) for fine-grained verification of candidate regions. Evaluated on the ANDH and ANDH-Full benchmarks, PSC-AVDN achieves a new state-of-the-art among zero-shot methods, matching or even surpassing the performance of several fine-tuned models.

Abstract

In this paper, we tackle the Aerial Vision-and-Dialog Navigation (AVDN) task in the training-free setting for resource-efficient high-altitude UAV navigation.Naively applying MLLMs leads to unreliable navigation due to weak directional grounding and the lack of explicit spatial memory.To address these issues, we propose PSC-AVDN, a training-free framework that tightly couples a three-stage Parsing-Search-Confirmation reasoning pipeline with a Structured Spatial Memory (SSM).The parsing stage uses an LLM to convert ambiguous dialogue instructions into stable geometric directional and destination cues.A Search Chain-of-Thought (S-CoT) then performs stepwise target exploration under high-altitude observations, and a Confirmation Chain-of-Thought (C-CoT) conducts fine-grained verification around candidate regions to resolve visual ambiguity.Meanwhile, SSM integrates three complementary sources of spatial cues, including multi-scale visual observation, spatial visual memory, and structured geometric memory to provide global spatial context and long-horizon consistency.Extensive experiments on ANDH and ANDH-Full show that PSC-AVDN establishes new state-of-the-art performance in the training-free setting, matching or surpassing several finetuned methods.Code will be publicly available at: https://github.com/QY6616/PSC-AVDN

Problem

• Aerial Vision-and-Dialog Navigation
• training-free
• spatial memory
• directional grounding
• UAV navigation

Innovation

• training-free navigation
• Chain-of-Thought reasoning
• Structured Spatial Memory
• Aerial Vision-and-Dialog Navigation
• geometric grounding