Skip to content

论文arXiv 2611.1303411 天前

RoboACE框架实现机器人对抗共进化,LIBERO-PRO成功率达74% ​

本文由 AI 根据公开内容整理,可能有误,请以原文为准:RoboACE: Adversarial Co-Evolution for Robot Self-Improvement(arXiv 2611.13034)

核心内容 ​

针对固定练习源导致机器人难以发现新弱点的问题,研究团队提出RoboACE(Robot Adversarial Co-Evolution)框架。该方法在冻结基础模型前提下,通过课程生成器与执行修正器的交替更新实现对抗共进化。在LIBERO系列四个面板测试中,该框架生成的冻结部署版本全面超越现有控制器与智能体系统,其中在LIBERO-PRO任务集上达到74.0%成功率,较初始部署提升15.0个百分点,比VLA控制器直接执行高出13.5个百分点。

技术架构包含课程线束与执行线束两大组件:前者利用累积能力证据动态生成高信息量挑战,后者响应并产出可复用的执行策略修订。系统采用交替固定更新机制,将已采纳的挑战固化为快照以对比执行修订效果。实验从紧凑的LIBERO练习条件集出发,验证了证据引导的练习选择、抽屉操作技能接地及控制器干预等关键模块的有效性,证明了在不微调大模型条件下通过结构化对抗训练提升具身智能泛化能力的可行性。

关键看点 ​

  • 对抗共进化机制:课程与执行线束交替更新,利用能力证据生成挑战并固化快照对比修订效果
  • LIBERO-PRO性能:冻结模型部署成功率达74.0%,较初始部署提升15.0%,超VLA直推13.5%
  • 免微调优化范式:在基础模型参数冻结前提下,仅通过外部课程与执行策略迭代实现自我改进
  • 可解释性验证:案例证实证据引导练习选择、抽屉技能接地及控制器干预机制的有效演进

工程与产业观察 ​

RoboACE验证了在冻结VLA模型下,通过对抗式课程学习与执行策略共进化可显著提升长尾任务成功率。这为具身智能提供了一条不依赖昂贵全量微调、利用运行时反馈持续挖掘模型潜力的工程路径,对降低机器人部署后的迭代成本具有重要参考价值。

英文摘要 (Abstract) ​

Robot agents can improve reusable skills, planning procedures, and recovery strategies from execution feedback, yet a fixed practice source leaves the search for new weaknesses unchanged. Useful practice depends on the agent's current capabilities, so the procedure that chooses practice must also be revisable. We present RoboACE (Robot Adversarial Co-Evolution), a framework that jointly evolves curriculum and execution harnesses around frozen foundation models. The curriculum harness uses accumulated capability evidence to propose informative challenges, while the execution harness responds with reusable revisions that expand execution capabilities. Alternating updates hold one side fixed as the other changes, and admitted challenges form fixed snapshots for comparing execution revisions. Starting from a compact set of LIBERO practice conditions, the resulting frozen deployment outperforms the evaluated controllers and agent systems across four LIBERO-family panels. On LIBERO-PRO, it reaches 74.0% success, exceeding the initial deployment by 15.0 percentage points and direct execution by its VLA controller by 13.5 points. Case studies document changes to evidence-guided practice selection, drawer-skill grounding, and controller intervention.

作者:Delin Zhao, Zhengrong Yue, Junlin He, Yali Wang

基于本页技术方向关联的整机参数、核心零部件、实测报告与同赛道选型横评。

外部来源

本页事实对应的公开页面,以原文为准。(点击将离开本站)

  1. arXiv 2611.13034:RoboACE: Adversarial Co-Evolution for Robot Self-Improvement ↗访问 2026-11-11