arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2412.05313cs.ROcs.AIcs.LG

λ：长时程室内移动操作机器人数据效率基准

λ: A Benchmark for Data-Efficiency in Long-Horizon Indoor Mobile Manipulation Robotics

Brown University（布朗大学）；Rutgers University（罗格斯大学）；University of Pennsylvania（宾夕法尼亚大学）

Ahmed Jaafar, Shreyas Sundara Raman, Sudarshan Harithas, Yichen Wei, Sofia Juliani, Anneke Wernerfelt, Benedict Quartey, Ifrah Idrees, Jason Xinyu Liu, Stefanie Tellex

备注：
Accepted to IROS 2025. Sudarshan Harithas and Yichen Wei contributed equally. 8 pages. 7 figures
阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对长时程移动操作任务数据效率低的问题，提出LAMBDA基准，包含571个人类演示，评估发现神经符号方法优于端到端学习方法且所需数据更少。

摘要：

学习执行长时程移动操作任务对于推动机器人在家庭和工作场所中的应用至关重要。然而，当前方法通常数据效率低下，这凸显了需要改进模型以及使用规模切合实际的基准来评估其效率的必要性。为解决这一问题，我们引入了LAMBDA（λ）基准——面向定向活动的长时程移动操作基准——该基准使用规模适中、更易于收集的数据集，评估模型在语言条件化、长时程、多房间、多楼层、抓取放置任务上的数据效率。我们的基准包含571个人类收集的演示，在模拟和真实世界环境中提供了真实性和多样性。与规划器生成的数据不同，这些轨迹提供了自然的变异性和可重放验证性，确保了稳健的学习和评估。我们利用λ来基准测试当前的端到端学习方法以及一种结合基础模型与任务和运动规划的模块化神经符号方法。我们发现，学习方法即使经过预训练，成功率也较低，而神经符号方法表现显著更好且所需数据更少。

英文摘要：

Learning to execute long-horizon mobile manipulation tasks is crucial for advancing robotics in household and workplace settings. However, current approaches are typically data-inefficient, underscoring the need for improved models that require realistically sized benchmarks to evaluate their efficiency. To address this, we introduce the LAMBDA (λ) benchmark-Long-horizon Actions for Mobile-manipulation Benchmarking of Directed Activities-which evaluates the data efficiency of models on language-conditioned, long-horizon, multi-room, multi-floor, pick-and-place tasks using a dataset of manageable size, more feasible for collection. Our benchmark includes 571 human-collected demonstrations that provide realism and diversity in simulated and real-world settings. Unlike planner-generated data, these trajectories offer natural variability and replay-verifiability, ensuring robust learning and evaluation. We leverage λ to benchmark current end-to-end learning methods and a modular neuro-symbolic approach that combines foundation models with task and motion planning. We find that learning methods, even when pretrained, yield lower success rates, while a neuro-symbolic method performs significantly better and requires less data.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑