CycleManip：通过有效历史感知与理解实现基于周期的操作
CycleManip: Enabling Cycle-based Manipulation via Effective History Perception and Understanding
面向视觉-语言-动作（VLA）模仿策略的即插即用历史感知模块：用代价感知采样挑选历史帧、用多任务学习增强历史理解，可直接嵌入现有 VLA 策略以处理需要精确终止时刻的周期性操作任务。
原文依据 the plug-and-play ability on imitation policies such as Vision-Language-Action (VLA) models
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论本文探索机器人操作中一个重要但尚未被充分研究的任务：基于周期的操作，即机器人需要在期望的终止时刻执行周期性或重复性动作。这类任务在日常生活中至关重要，例如摇晃瓶子或敲钉子。然而，此前鲜有工作探索该任务，由此带来两个主要挑战：1）由于对历史信息的利用不充分，模仿方法往往无法在期望的终止时刻内完成任务；2）缺乏具备充足数据和自动评测工具的基准，阻碍了该领域有效方案的发展。为应对这些挑战，我们首先提出 CycleManip 框架，以端到端的模仿方式实现基于周期的任务操作，且无需任何额外模型、层次化结构或显著的计算开销。其核心思路是通过代价感知的采样策略增强有效的历史感知，并通过多任务学习提升对历史的理解。其次，我们提出了一个基于周期的任务操作基准，提供多样化的周期任务以及一种自动评测方法。在仿真和真实环境中开展的大量实验表明，我们的方法在基于周期的任务操作中取得了高成功率。结果进一步显示，该方法在通用操作任务中具有较强的适应性能，并且对视觉-语言-动作（VLA）模型等模仿策略具备即插即用的能力。此外，结果表明我们的方法可应用于多种机器人平台，包括双臂夹爪、灵巧手和人形机器人。
英文原文摘要 1693 字符
In this paper, we explore an important yet underexplored task in robot manipulation: cycle-based manipulation, where robots need to perform cyclic or repetitive actions with an expected terminal time. These tasks are crucial in daily life, such as shaking a bottle or knocking a nail. However, few prior works have explored this task, leading to two main challenges: 1) the imitation methods often fail to complete these tasks within the expected terminal time due to the ineffective utilization of history; 2) the absence of a benchmark with sufficient data and automatic evaluation tools hinders development of effective solutions in this area. To address these challenges, we first propose the CycleManip framework to achieve cycle-based task manipulation in an end-to-end imitation manner without requiring any extra models, hierarchical structure or significant computational overhead. The core insight is to enhance effective history perception by a cost-aware sampling strategy and to improve historical understanding by multi-task learning. Second, we introduce a cycle-based task manipulation benchmark, which provides diverse cycle-based tasks, and an automatic evaluation method. Extensive experiments conducted in both simulation and real-world settings demonstrate that our method achieves high success rates in cycle-based task manipulation. The results further show strong adaptability performance in general manipulation, and the plug-and-play ability on imitation policies such as Vision-Language-Action (VLA) models. Moreover, the results show that our approach can be applied across diverse robotic platforms, including bi-arm grippers, dexterous hands, and humanoid robots.