Yggdrasil：一种面向实时查询的层优先三维场景图

中文摘要 AI 辅助

机器人代理使用三维场景图（3DSG）来执行从场景理解到场景交互的各种任务。尽管已有大量工作致力于场景图生成，但很少关注优化图的消费，这使得最先进的感知流水线不得不围绕自己的场景图工作，并承担与感知循环实时预算不符的延迟成本。我们提出了Yggdrasil，这是第一个在生成和消费两方面都高效的3D场景图：一种基于通用节点、边和层的层优先分层图，它表达了现有流水线已经产生的表示，无论是室内还是室外、平面还是分层，同时原生回答下游任务发出的位置和语义查询。与已发表的DSG基线相比，Yggdrasil回答查询的速度最高可达121倍，我们测量的每个查询耗时在2到127微秒之间，比3DSG消费者所处的200微秒关键帧预算快三到五个数量级，在工作站和嵌入式设备上均如此。我们将Yggdrasil集成到三个已发表的流水线中，涵盖人类轨迹预测、目标导航和人类感知运动规划，在这些流水线中，它消除了每个流水线在场景图上花费的时间高达99%。该实现、基准测试工具以及所有三个集成均可在线获取。

英文摘要

Robotic agents use 3D scene graphs (3DSG) to perform tasks ranging from scene understanding to scene interaction. Although an extensive body of work addresses scene graph generation, little attention has been paid to optimizing the graph for consumption, which leaves state-of-the-art perception pipelines to work around their own scene graph and to pay a latency cost that does not fit the real-time budget a perception loop runs on. We present Yggdrasil, the first 3D scene graph designed to be efficient for both generation and consumption: a layer-first hierarchical graph built from generic nodes, edges, and layers, which expresses the representations existing pipelines already produce, indoor or outdoor, flat or hierarchical, while natively answering the positional and semantic queries downstream tasks issue. Against a published DSG baseline, Yggdrasil answers queries up to $121\times$ faster, and every query we measure falls between 2 and 127 microseconds, three to five orders of magnitude inside the 200 microsecond keyframe budget a 3DSG consumer lives in, on both a workstation and embedded class device. We integrate Yggdrasil into three published pipelines spanning human trajectory prediction, object-goal navigation, and human-aware motion planning, where it removes up to 99% of the time each spends on its scene graph. The implementation, benchmark harness, and all three integrations are available online.

发表机构

• San Diego State University（圣地亚哥州立大学）
• KTH Royal Institute of Technology（瑞典皇家理工学院）
• University of Illinois Chicago（伊利诺伊大学芝加哥分校）

机构由 AI 辅助整理，请以论文原文为准。