研究人员开发了 DynoSLAM，一种新颖的动态 GraphSLAM 架构，它将图神经网络 (GNN) 集成到因子图优化中，以改善在拥挤环境中的机器人导航。该系统将行人运动预测建模为随机世界模型，使用来自训练有素的 GNN 的蒙特卡洛滚动来捕捉人类互动的不确定性。该方法将这种不确定性嵌入到 SLAM 图中，从而实现更准确的跟踪并防止优化失败，最终为无碰撞机器人导航提供概率安全包络。 AI
影响 通过改进定位和安全预测，增强了在动态、人口稠密空间中的机器人导航能力。
排序理由 学术论文，详细介绍了使用 GNN 的一种新的 SLAM 方法。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
arXiv:2605.02759v1 Announce Type: cross Abstract: Traditional Simultaneous Localization and Mapping (SLAM) algorithms rely heavily on the static environment assumption, which severely limits their applicability in real-world spaces populated by moving entities, such as pedestrian…
Traditional Simultaneous Localization and Mapping (SLAM) algorithms rely heavily on the static environment assumption, which severely limits their applicability in real-world spaces populated by moving entities, such as pedestrians. In this work, we propose DynoSLAM, a tightly-co…