解析、搜索与确认：基于思维链推理与结构化空间记忆的免训练航拍视觉对话导航
Parse, Search, and Confirmation: Training-Free Aerial Vision-and-Dialog Navigation with Chain-of-Thought Reasoning and Structured Spatial Memory
把免训练 MLLM 的导航能力拆成「解析指令—搜索探索—确认验证」三段思维链，并配一个融合多尺度观测、视觉记忆与几何记忆的结构化空间记忆，可直接迁移到任何需要长时程空间推理的视觉语言任务（如视觉定位、具身导航）。
原文依据 a three-stage Parsing-Search-Confirmation reasoning pipeline with a Structured Spatial Memory (SSM)
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论本文研究免训练设定下的航拍视觉对话导航（AVDN）任务，以实现资源高效的高空无人机导航。直接套用多模态大模型会因方向视觉定位能力薄弱以及缺乏显式空间记忆而导致导航不可靠。为解决上述问题，本文提出 PSC-AVDN，一个免训练框架，将三阶段的「解析—搜索—确认」推理流水线与结构化空间记忆（SSM）紧密耦合。解析阶段利用大语言模型将含糊的对话指令转换为稳定的几何方向线索与目的地线索；随后，搜索思维链（S-CoT）在高空观测下执行逐步的目标探索，确认思维链（C-CoT）则在候选区域周围进行细粒度验证，以消解视觉歧义。与此同时，SSM 融合三类互补的空间线索，包括多尺度视觉观测、空间视觉记忆与结构化几何记忆，从而提供全局空间上下文与长时程一致性。在 ANDH 与 ANDH-Full 上的大量实验表明，PSC-AVDN 在免训练设定下取得了新的最优性能，与若干微调方法持平甚至超越。
英文原文摘要 1267 字符
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.