研究人员开发了两个新的空中视觉语言导航（VLN）框架。DreamFly基于Dream-VLA，采用因果记忆和递减视界扩散规划方法来提高在复杂环境中的导航能力。而RecoverFly则采用故障感知强化学习后训练框架，以提高端到端UAV-VLA策略的性能和泛化能力。这两种方法在基准数据集上都取得了显著的改进，在各种指标上都优于现有方法。 AI
影响 空中导航的这些进步可能带来更强大的自主无人机，用于检查、交付和监视等任务。
排序理由 该集群包含两篇详细介绍新的空中视觉语言导航框架的研究论文。
在 Hugging Face Daily Papers 阅读 →
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
完整方法见我们的编辑标准。
arXiv:2608.12308v1 Announce Type: cross Abstract: Aerial vision-language navigation (VLN) requires an embodied agent to integrate visual evidence over time, plan future actions, and determine when it has reached a navigation goal under partial observability. Although recent VLA m…
Aerial vision-language navigation (VLN) requires an embodied agent to integrate visual evidence over time, plan future actions, and determine when it has reached a navigation goal under partial observability. Although recent VLA models offer a promising perception-to-action parad…
arXiv:2608.09467v1 Announce Type: cross Abstract: Unmanned aerial vehicle vision-language navigation (UAV-VLN) requires agents to translate visual observations and language instructions into reliable flight actions in complex environments. Although recent end-to-end UAV vision-la…
Unmanned aerial vehicle vision-language navigation (UAV-VLN) requires agents to translate visual observations and language instructions into reliable flight actions in complex environments. Although recent end-to-end UAV vision-language-action (UAV-VLA) policies reduce reliance o…