通过时空剪枝提升视觉语言导航效率，无需重训练即可实现实时部署。
History-Conditioned Spatio-Temporal Visual Token Pruning for Efficient Vision-Language Navigation
- 基于注意力与查询引导的时空裁剪，保留关键视觉信息。
- 极端剪枝下仍保持高导航准确率，推理速度显著提升。
- 可直接接入现有系统，适合实际机器人实时导航场景。
视觉语言导航（VLN）使机器人能在视觉语境中理解自然语言指令，是具身机器人系统的关键能力。近期的视觉-语言-动作（VLA）模型虽表现优异，但计算开销大，导致延迟高，难以实现实时部署。本文提出一种无需训练的时空视觉标记剪枝框架，针对基于VLA的VLN任务。通过当前视图的空间标记选择，结合历史记忆的时空压缩，实现长周期推理的高效性，减少冗余计算。利用基于注意力的标记重要性评估和查询引导的时空过滤机制，该方法在不重新训练或修改预训练模型的前提下，保留导航相关关键信息，支持即插即用集成到现有VLA系统中。在标准VLN基准测试中，本方法显著优于现有剪枝策略，在极端剪枝条件下仍保持优异导航准确率，同时具备极高的推理效率。在Unitree Go2四足机器人上的真实世界部署进一步验证了其在实际机器人约束下的可靠低延迟指令跟随能力。期望此工作能弥合大规模多模态建模与实际机器人实时部署之间的差距。
原文摘要 · Abstract (English)
Vision-Language Navigation (VLN) enables robots to follow natural-language instructions in visually grounded environments, serving as a key capability for embodied robotic systems. Recent Vision-Language-Action (VLA) models have demonstrated strong navigation performance, but their high computational cost introduces latency that limits real-time deployment. We propose a training-free spatio-temporal vision token pruning framework tailored to VLA-based VLN. We apply spatial token selection to the current view, alongside spatio-temporal compression for historical memories, enabling efficient long-horizon inference while reducing redundant computation. Leveraging attention-based token importance and query-guided spatio-temporal filtering, the proposed approach preserves navigation-relevant information without retraining or modifying pretrained models, allowing plug-and-play integration into existing VLA systems. Through experiments on standard VLN benchmarks, we confirm that our method significantly outperforms existing pruning strategies. It successfully preserves superior navigation accuracy under extreme pruning scenarios, all while maintaining the highly competitive inference efficiency. Real-world deployment on a Unitree Go2 quadruped robot further validates reliable and low-latency instruction-following navigation under practical robotic constraints. We hope this work helps bridge the gap between large-scale multimodal modeling and efficient, real-time embodied deployment in robotic navigation systems. Project Page: https://wqtwjt1996.github.io/publications/2026-vln.html
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。