迈向人在回路的机器人故障恢复：弥合人机协作中的沟通差距
Toward Human-in-the-Loop Robot Failure Recovery: Bridging Communication Gaps in Human-Robot Collaboration
- Cornell University（康奈尔大学）
- Cornell Tech（康奈尔科技校区）
- Brown University（布朗大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
本文提出LD-HRI基准，通过游戏和数据集评估说话者请求，发现LLM请求虽提高新手成功率，但专家-新手差距仍显著，为弥合人机协作沟通差距提供基础。
AI中文摘要：
机器人可以通过向旁观者求助来从故障中恢复，但有效的人在回路恢复需要考虑到人们知识差异的沟通。先前的逆语义学工作使用单一听者模型生成请求，未测试听者知识差异的影响。我们引入了人机交互中的听者差异（LD-HRI），这是一个通过人类听者表现来评估说话者的游戏、数据集和基准。我们的评估在受控的听者信息差异下，考察了请求属性、大语言模型（LLM）说话者以及逆语义学请求选择算法。该语料库包含446条人类编写的请求和1,302次听者试验。我们还评估了24条冻结的LLM编写的请求，涉及70名人类听者，共560次试验。在全部四项任务中，新手成功率在描述性上更高，使用模型编写的请求，但两种请求来源都留下了显著的专家-新手差距，包括LLM请求的16个百分点。LD-HRI使这些差距可测量，为设计更稳健的人机和人机交互沟通提供了基础。
英文摘要：
Robots can recover from failures by asking bystanders for help, but effective human-in-the-loop recovery requires communication that accounts for differences in people's knowledge. Prior inverse-semantics work generates requests using a single listener model, leaving differences in listener knowledge untested. We introduce Listener Differences in Human-Robot Interaction (LD-HRI), a game, dataset, and benchmark that evaluates speakers through human listener performance. Our evaluation examines request properties, large language model (LLM) speakers, and inverse-semantics request-selection algorithms under controlled differences in listener information. The corpus contains 446 human-written requests and 1{,}302 listener trials. We additionally evaluated 24 frozen LLM-written requests with 70 human listeners across 560 trials. Novice success is descriptively higher with model-written requests across all four tasks, yet both request sources leave substantial expert--novice gaps, including 16 percentage points for LLM requests. LD-HRI makes these gaps measurable, providing a foundation for designing more robust communication in human-robot and human-agent interaction.
