Research

Paper

AI LLM March 18, 2026

AgentVLN: Towards Agentic Vision-and-Language Navigation

Authors

Zihao Xin, Wentong Li, Yixuan Jiang, Ziyuan Huang, Bin Wang, Piji Li, Jianke Zhu, Jie Qin, Shengjun Huang

Abstract

Vision-and-Language Navigation (VLN) requires an embodied agent to ground complex natural-language instructions into long-horizon navigation in unseen environments. While Vision-Language Models (VLMs) offer strong 2D semantic understanding, current VLN systems remain constrained by limited spatial perception, 2D-3D representation mismatch, and monocular scale ambiguity. In this paper, we propose AgentVLN, a novel and efficient embodied navigation framework that can be deployed on edge computing platforms. We formulate VLN as a Partially Observable Semi-Markov Decision Process (POSMDP) and introduce a VLM-as-Brain paradigm that decouples high-level semantic reasoning from perception and planning via a plug-and-play skill library. To resolve multi-level representation inconsistency, we design a cross-space representation mapping that projects perception-layer 3D topological waypoints into the image plane, yielding pixel-aligned visual prompts for the VLM. Building on this bridge, we integrate a context-aware self-correction and active exploration strategy to recover from occlusions and suppress error accumulation over long trajectories. To further address the spatial ambiguity of instructions in unstructured environments, we propose a Query-Driven Perceptual Chain-of-Thought (QD-PCoT) scheme, enabling the agent with the metacognitive ability to actively seek geometric depth information. Finally, we construct AgentVLN-Instruct, a large-scale instruction-tuning dataset with dynamic stage routing conditioned on target visibility. Extensive experiments show that AgentVLN consistently outperforms prior state-of-the-art methods (SOTA) on long-horizon VLN benchmarks, offering a practical paradigm for lightweight deployment of next-generation embodied navigation models. Code: https://github.com/Allenxinn/AgentVLN.

Metadata

arXiv ID: 2603.17670
Provider: ARXIV
Primary Category: cs.RO
Published: 2026-03-18
Fetched: 2026-03-19 06:01

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.17670v1</id>\n    <title>AgentVLN: Towards Agentic Vision-and-Language Navigation</title>\n    <updated>2026-03-18T12:43:47Z</updated>\n    <link href='https://arxiv.org/abs/2603.17670v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.17670v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Vision-and-Language Navigation (VLN) requires an embodied agent to ground complex natural-language instructions into long-horizon navigation in unseen environments. While Vision-Language Models (VLMs) offer strong 2D semantic understanding, current VLN systems remain constrained by limited spatial perception, 2D-3D representation mismatch, and monocular scale ambiguity. In this paper, we propose AgentVLN, a novel and efficient embodied navigation framework that can be deployed on edge computing platforms. We formulate VLN as a Partially Observable Semi-Markov Decision Process (POSMDP) and introduce a VLM-as-Brain paradigm that decouples high-level semantic reasoning from perception and planning via a plug-and-play skill library. To resolve multi-level representation inconsistency, we design a cross-space representation mapping that projects perception-layer 3D topological waypoints into the image plane, yielding pixel-aligned visual prompts for the VLM. Building on this bridge, we integrate a context-aware self-correction and active exploration strategy to recover from occlusions and suppress error accumulation over long trajectories. To further address the spatial ambiguity of instructions in unstructured environments, we propose a Query-Driven Perceptual Chain-of-Thought (QD-PCoT) scheme, enabling the agent with the metacognitive ability to actively seek geometric depth information. Finally, we construct AgentVLN-Instruct, a large-scale instruction-tuning dataset with dynamic stage routing conditioned on target visibility. Extensive experiments show that AgentVLN consistently outperforms prior state-of-the-art methods (SOTA) on long-horizon VLN benchmarks, offering a practical paradigm for lightweight deployment of next-generation embodied navigation models. Code: https://github.com/Allenxinn/AgentVLN.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.RO'/>\n    <published>2026-03-18T12:43:47Z</published>\n    <arxiv:comment>19pages, 4 figures</arxiv:comment>\n    <arxiv:primary_category term='cs.RO'/>\n    <author>\n      <name>Zihao Xin</name>\n    </author>\n    <author>\n      <name>Wentong Li</name>\n    </author>\n    <author>\n      <name>Yixuan Jiang</name>\n    </author>\n    <author>\n      <name>Ziyuan Huang</name>\n    </author>\n    <author>\n      <name>Bin Wang</name>\n    </author>\n    <author>\n      <name>Piji Li</name>\n    </author>\n    <author>\n      <name>Jianke Zhu</name>\n    </author>\n    <author>\n      <name>Jie Qin</name>\n    </author>\n    <author>\n      <name>Shengjun Huang</name>\n    </author>\n  </entry>"
}