arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2512.17853cs.ROcs.AI
收藏

AnyTask: 一种用于推进仿真到现实策略学习的任务和数据生成框架

AnyTask: an Automated Task and Data Generation Framework for Advancing Sim-to-Real Policy Learning

Robotics and AI Institute(机器人与人工智能研究所)
Ran Gong, Xiaohan Zhang, Jinghuan Shang, Maria Vittoria Minniti, Jigarkumar Patel, Valerio Pepe, Riedana Yan, Ahmet Gundogdu, Ivan Kapelyukh, Ali Abbas, Xiaoqia… 展开作者

Ran Gong, Xiaohan Zhang, Jinghuan Shang, Maria Vittoria Minniti, Jigarkumar Patel, Valerio Pepe, Riedana Yan, Ahmet Gundogdu, Ivan Kapelyukh, Ali Abbas, Xiaoqiang Yan, Harsh Patel, Laura Herlant, Karl Schmeckpeper

AI总结:

AnyTask通过自动化框架结合GPU仿真与基础模型,生成多样化任务和机器人数据,提升仿真到现实策略学习的效率与效果。

摘要:

通用机器人学习仍然受到数据的限制:大规模、多样性和高质量的交互数据在现实世界中收集成本高昂。尽管仿真已成为扩大数据收集的一种有希望的方法,但相关的任务,包括仿真任务设计、任务感知场景生成、专家示范合成以及仿真到现实转移,仍然需要大量的人力工作。我们提出了AnyTask,一种自动化框架,将大规模并行GPU仿真与基础模型相结合,以设计多样化的操作任务并合成机器人数据。我们介绍了三个AnyTask代理,旨在解决尽可能多的任务:1)ViPR,一种新颖的任务和运动规划代理,具有VLM-in-the-loop平行细化;2)ViPR-Eureka,一种强化学习代理,具有生成的密集奖励和LLM引导的接触采样;3)ViPR-RL,一种混合规划和学习方法,仅使用稀疏奖励共同生成高质量的示范。我们将在生成的数据上训练行为克隆策略,验证它们在仿真中的表现,并直接部署在现实机器人硬件上。这些策略能够推广到新的物体姿态,在一系列现实世界的拾取和放置、抽屉开启、接触丰富的推动以及长周期操作任务中,平均成功率为44%。我们的项目网站位于https://anytask.rai-inst.com。

英文摘要:

Generalist robot learning remains constrained by data: large-scale, diverse, and high-quality interaction data are expensive to collect in the real world. While simulation has become a promising way for scaling up data collection, the related tasks, including simulation task design, task-aware scene generation, expert demonstration synthesis, and sim-to-real transfer, still demand substantial human effort. We present AnyTask, an automated framework that pairs massively parallel GPU simulation with foundation models to design diverse manipulation tasks and synthesize robot data. We introduce three AnyTask agents for generating expert demonstrations aiming to solve as many tasks as possible: 1) ViPR, a novel task and motion planning agent with VLM-in-the-loop Parallel Refinement; 2) ViPR-Eureka, a reinforcement learning agent with generated dense rewards and LLM-guided contact sampling; 3) ViPR-RL, a hybrid planning and learning approach that jointly produces high-quality demonstrations with only sparse rewards. We train behavior cloning policies on generated data, validate them in simulation, and deploy them directly on real robot hardware. The policies generalize to novel object poses, achieving 44% average success across a suite of real-world pick-and-place, drawer opening, contact-rich pushing, and long-horizon manipulation tasks. Our project website is at https://anytask.rai-inst.com .

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。

↑