DecoVLN: Decoupling Observation, Reasoning, and Correction for Vision-and-Language Navigation
2026-03-13 • Robotics
Robotics
AI summaryⓘ
The authors tackle Vision-and-Language Navigation (VLN), where agents follow long instructions to move in 3D spaces. They identify two main problems: building a useful memory of past observations and fixing errors that pile up during navigation. Their solution, DecoVLN, creates a smart memory by carefully picking important frames based on how relevant, diverse, and well-distributed they are. They also improve error correction by using state-action pairs filtered by how far the agent strays from the ideal path. Their method works well in experiments and has even been tested in real-world scenarios.
Vision-and-Language NavigationLong-term MemoryClosed-loop ControlStreaming PerceptionSemantic RelevanceVisual DiversityTemporal CoverageCompounding ErrorsState-action PairsGeodesic Distance
Authors
Zihao Xin, Wentong Li, Yixuan Jiang, Bin Wang, Runming Cong, Jie Qin, Shengjun Huang
Abstract
Vision-and-Language Navigation (VLN) requires agents to follow long-horizon instructions and navigate complex 3D environments. However, existing approaches face two major challenges: constructing an effective long-term memory bank and overcoming the compounding errors problem. To address these issues, we propose DecoVLN, an effective framework designed for robust streaming perception and closed-loop control in long-horizon navigation. First, we formulate long-term memory construction as an optimization problem and introduce adaptive refinement mechanism that selects frames from a historical candidate pool by iteratively optimizing a unified scoring function. This function jointly balances three key criteria: semantic relevance to the instruction, visual diversity from the selected memory, and temporal coverage of the historical trajectory. Second, to alleviate compounding errors, we introduce a state-action pair-level corrective finetuning strategy. By leveraging geodesic distance between states to precisely quantify deviation from the expert trajectory, the agent collects high-quality state-action pairs in the trusted region while filtering out the polluted data with low relevance. This improves both the efficiency and stability of error correction. Extensive experiments demonstrate the effectiveness of DecoVLN, and we have deployed it in real-world environments.