面向长程重排任务的主动推理移动操作
Mobile Manipulation with Active Inference for Long-Horizon Rearrangement Tasks
AI总结:
针对主动推理难以用于复杂长程机器人任务的问题,提出分层主动推理架构,结合高层技能选择与全身控制,在 Habitat Benchmark 三项任务上优于现有基线。
摘要:
尽管主动推理在机器人控制中的兴趣不断增长,但其在复杂、长程任务中的应用仍未得到检验。我们通过引入一种完全分层的主动推理架构来弥补这一空白,该架构用于现实机器人环境中的目标导向行为。我们的模型结合了一个高层主动推理模型,用于在由全身主动推理控制器实现的离散技能之间进行选择。这种统一方法无需离线训练即可实现灵活的技能组合、在线适应性以及任务失败后的恢复。在用于移动操作的 Habitat Benchmark 上进行评估时,我们的方法在三项长程任务上均优于最先进的基线方法,首次证明主动推理能够扩展到现代机器人基准测试的复杂程度。
英文摘要:
Despite growing interest in active inference for robotic control, its application to complex, long-horizon tasks remains untested. We address this gap by introducing a fully hierarchical active inference architecture for goal-directed behavior in realistic robotic settings. Our model combines a high-level active inference model that selects among discrete skills realized via a whole-body active inference controller. This unified approach enables flexible skill composition, online adaptability, and recovery from task failures without requiring offline training. Evaluated on the Habitat Benchmark for mobile manipulation, our method outperforms state-of-the-art baselines across the three long-horizon tasks, demonstrating for the first time that active inference can scale to the complexity of modern robotics benchmarks.
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。