Navi-Agent: Unlocalized Monocular Navigation Agent

2026-09-17

AI Summary

本文提出Navi-Agent，通过构建基于视觉观察和运动历史的无坐标空间状态来解决零样本VLN-CE中的持续空间意识维护问题。

Abstract

Vision-Language Navigation in Continuous Environments (VLN-CE) requires an embodied agent to execute long-horizon instructions in unknown environments. Existing zero-shot VLN-CE systems typically maintain spatial states through geometric localization or coordinate-based representations. Recent geometry-constrained navigation removes depth and globally consistent coordinates, but maintaining persistent spatial awareness for place confirmation, progress verification, and recovery remains challenging. We present Navi-Agent, a zero-shot VLN-CE agent that constructs a coordinate-free spatial state from visual observations and executed motion histories. Navi-Agent organizes this state as a navigation topology, where nodes represent visual places and edges represent motion transitions. This representation enables observation-based approximate self-localization, task progress verification, and visual revisitation-based recovery. Navi-Agent performs closed-loop navigation by decomposing instructions into sub-goals, executing local visual navigation, and verifying visited places through the constructed spatial state. Experiments on zero-shot VLN-CE benchmark and real-world robot platforms show that Navi-Agent achieves state-of-the-art performance among geometry-constrained methods while remaining competitive with approaches relying on geometric localization.

Problem

• Vision-Language Navigation
• Continuous Environments
• Zero-shot VLN-CE
• Spatial Awareness
• Geometry-Constrained Navigation

Innovation

• zero-shot VLN-CE
• coordinate-free spatial state
• navigation topology
• visual revisitation-based recovery