面向开放环境与指令：通过快慢交互推理的通用视觉语言导航
Towards Open Environments and Instructions: General Vision-Language Navigation via Fast-Slow Interactive Reasoning
提出快慢交互推理框架：快模块用端到端策略网络实时输出动作并积累历史记忆，慢模块对记忆做深度反思提取结构化经验，持续优化快模块，可直接迁移到视觉导航、具身决策等需要在线适应未见环境的视觉任务。
原文依据 we propose the slow4fast-VLN, establishing a dynamic interactive fast-slow reasoning framework
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论视觉语言导航（VLN）旨在使智能体根据语言指令导航至目标位置。传统VLN通常遵循闭集假设，即训练与测试数据共享相同风格的输入图像与指令。然而，现实世界是开放的，充满各种未见环境，这给闭集方法带来巨大困难。为此，我们聚焦通用场景适应（GSA-VLN）任务，旨在通过引入多样环境与不一致指令来学习泛化的导航能力。针对该任务，面对未见环境与指令时，挑战主要在于如何使智能体在导航过程中动态产生泛化策略。近期研究表明，借助快慢认知系统，人类能够生成稳定的策略，从而增强其对开放世界的适应能力。受此启发，我们提出slow4fast-VLN，建立动态交互的快慢推理框架。快推理模块是一个端到端策略网络，通过实时输入输出动作，并在历史仓库中累积执行记录以构建记忆。慢推理模块分析快推理模块生成的记忆，通过深度反思提取增强决策泛化能力的经验。这些经验以结构化方式存储，并用于持续优化快推理模块。与传统将快慢推理视为独立机制的方法不同，我们的框架实现了快慢交互。通过利用慢推理的经验，它持续提升快决策的准确性与泛化能力。这种交互使系统在面对未见场景时能够持续适应并高效执行导航任务。大量实验证明了我们方法的优越性。
英文原文摘要 1986 字符
Vision-Language Navigation (VLN) aims to enable agents to navigate to a target location based on language instructions. Traditional VLN often follows a close-set assumption, i.e., training and test data share the same style of the input images and instructions. However, the real world is open and filled with various unseen environments, posing enormous difficulties for close-set methods.To this end, we focus on the General Scene Adaptation (GSA-VLN) task, aiming to learn generalized navigation ability by introducing diverse environments and inconsistent intructions.Towards this task, when facing unseen environments and instructions, the challenge mainly lies in how to enable the agent to dynamically produce generalized strategies during the navigation process. Recent research indicates that by means of fast and slow cognition systems, human beings could generate stable policies, which strengthen their adaptation for open world. Inspired by this idea, we propose the slow4fast-VLN, establishing a dynamic interactive fast-slow reasoning framework.The fast-reasoning module, an end-to-end strategy network, outputs actions via real-time input. It accumulates execution records in a history repository to build memory.The slow-reasoning module analyze the memories generated by the fast-reasoning module. Through deep reflection, it extracts experiences that enhance the generalization ability of decision-making. These experiences are structurally stored and used to continuously optimize the fast-reasoning module. Unlike traditional methods that treat fast-slow reasoning as independent mechanisms, our framework enables fast-slow interaction. By leveraging the experiences from slow reasoning, it continually improves the accuracy and generalization ability of fast decisions. This interaction allows the system to continuously adapt and efficiently execute navigation tasks when facing unseen scenarios. Extensive experiments demonstrate the superiorities of our method.