论文arXiv 2611.07493今天

移动操作机器人开放词汇搜索新架构：实现实时可靠缺席检测

针对移动机器人物体搜索误报率高痛点，提出解耦检索与验证的两阶段架构。在ScanNet及真机测试中，以低一个数量级的检测器调用量匹配穷举搜索可靠性，单消费级GPU即可实现板载实时运行。

AI 整理

感知与导航
具身模型
端侧部署

本文由 AI 根据公开内容整理，可能有误，请以原文为准：Fast and Reliable Open-Vocabulary Object Search with Absence Detection from Robot Memory（arXiv 2611.07493）

核心内容

针对家庭护理等场景中移动操作机器人难以确认目标物体缺失的问题，研究指出现有文本或视觉嵌入搜索系统的特异性（Specificity）最高仅9%，极易对不存在物体产生幻觉。为此，论文提出一种两阶段架构，将快速候选检索与验证选择解耦，旨在利用带位姿的序列RGB观测数据，同时实现高召回率与可靠的缺席检测，解决传统方法依赖昂贵3D表示或处理缓慢的瓶颈。

该方案在ScanNet公寓场景及物理机器人上验证了8种检索-验证配置。结果显示，在基于字幕的检索后增加视觉验证步骤，其可靠性匹配甚至超过穷举搜索基线，且检测器调用量降低一个数量级。系统完全基于板载计算，仅需单张消费级GPU即可在真机上实现实时推理，证明了轻量级2D感知方案在开放词汇搜索任务中的工程可行性。

关键看点

缺席检测瓶颈：现有开放词汇搜索系统特异性最高仅9%，几乎无法正确判断目标物体缺失，导致无效导航与交互。
两阶段解耦架构：分离快速候选检索与视觉验证选择，利用带位姿序列RGB数据，兼顾高召回率与可靠缺席检测能力。
计算效率跃升：引入视觉验证后的字幕检索方案，检测器调用量较穷举搜索降低一个数量级，且保持同等可靠性。
端侧实时部署：算法在物理机器人单消费级GPU上实现全板载实时运行，摆脱对云端算力或昂贵3D传感器的依赖。

工程与产业观察

该研究解决了具身智能“确认物体不在场”的工程难题，通过架构解耦将开放词汇搜索从纯语义匹配升级为可验证的感知闭环。其单消费级GPU实时运行的特性，为服务机器人在非结构化环境中的低成本、高可靠部署提供了关键算法支撑，显著降低了端侧算力门槛。

英文摘要 (Abstract)

Mobile manipulators deployed in everyday human environments, such as homes for assistive care or flexible automation, often need to locate specific objects on demand from accumulated observations. Equally important, they must reliably determine when a target is absent before committing to navigation or interaction. Existing approaches demand either slow exhaustive re-processing or expensive 3D representations, and are evaluated on recall and accuracy rather than specificity, leaving their ability to abstain on missing targets unmeasured. We find existing systems that search using text or visual embedding almost never abstain on absent objects achieving at most 9% specificity. To support accurate search at speed, we present a two-stage architecture that decouples fast candidate retrieval from verified selection, enabling both high recall and reliable absence detection from sequential RGB observations with pose. We evaluate eight retrieval-selection configurations plus an exhaustive search baseline across ScanNet apartment scenes, including significant numbers of absent objects. Adding a visual verification step to caption-based retrieval matches or exceeds exhaustive search reliability while requiring an order of magnitude fewer detector calls. We reproduce ScanNet scores on a physical robot operating entirely on-board with a single consumer GPU running in real time.

作者：Nicholas Paul, Devson Butani, Eric Martinson