全天多场景终身视觉-语言导航与Tucker适应
All-day Multi-scenes Lifelong Vision-and-Language Navigation with Tucker Adaptation
AI总结：
本文提出Tucker适应方法，解决视觉-语言导航在多场景下的持续学习问题，通过高阶张量分解实现知识解耦，提升长期部署灵活性。
摘要：
部署视觉-语言导航（VLN）代理需要在多样场景和环境中进行适应，但针对特定场景的微调常导致其他场景的知识遗忘，限制了长期灵活部署。本文将这一挑战形式化为全天多场景终身VLN（AML-VLN）问题。现有参数高效适配器（如LoRA及其变种）受限于二维矩阵形式，无法捕捉跨多场景和环境的多级导航知识。为此，我们提出Tucker适应（TuKA），将多级导航知识表示为高阶张量，并利用Tucker分解将知识解耦为共享子空间和场景特定专家。我们进一步引入解耦知识增量学习策略，以巩固共享子空间并约束特定专家，实现解耦的终身学习。基于TuKA，我们还开发了名为AlldayWalker的VLN代理，能够在多个导航场景中持续学习，实现全天多场景导航。大量实验表明，AlldayWalker在性能上始终优于现有最先进基线。
英文摘要：
Deploying vision-and-language navigation (VLN) agents requires adaptation across diverse scenes and environments, but fine-tuning on a specific scenario often causes catastrophic forgetting in others, which severely limits flexible long-term deployment. We formalize this challenge as the all-day multi-scenes lifelong VLN (AML-VLN) problem. Existing parameter-efficient adapters (e.g., LoRA and its variants) are limited by their two-dimensional matrix form, which fails to capture the multi-hierarchical navigation knowledge spanning multiple scenes and environments. To address this, we propose Tucker Adaptation (TuKA), which represents the multi-hierarchical navigation knowledge as a high-order tensor and leverages Tucker decomposition to decouple the knowledge into shared subspaces and scenario-specific experts. We further introduce a decoupled knowledge incremental learning strategy to consolidate shared subspaces while constraining specific experts for decoupled lifelong learning. Building on TuKA, we also develop a VLN agent named AlldayWalker, which continually learns across multiple navigation scenarios, achieving all-day multi-scenes navigation. Extensive experiments show that AlldayWalker consistently outperforms state-of-the-art baselines.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
