Skip to content

论文arXiv 2611.176586 天前

模仿策略习得本体锚点:感知噪声折扣预测律与闭环代价 ​

本文由 AI 根据公开内容整理,可能有误,请以原文为准:Imitation Policies Learn a Proprioceptive Anchor: A Predictive Law for Perception Noise and Its Closed Loop Cost(arXiv 2611.17658)

核心内容 ​

针对机器人操作模仿学习中感知噪声处理机制,研究提出“本体锚点”理论并推导出感知读数折扣权重公式 $\alpha = V/(V + s^2)$。在RoboMimic仿真任务中,该模型对10项保持测试灵敏度的预测平均绝对误差仅为0.025,而经典变量误差(EIV)回归高达0.37。参数无关偏差检验排除了策略单纯高估噪声的假设,确立了本体状态作为信息源的有效性。

研究发现训练期的噪声折扣并未完全迁移至闭环部署阶段,15个通过注册门槛的策略对持续读数偏移的敏感度均超过理论 $\alpha$ 值。前瞻性测试确认低 $\alpha$ 导致Diffusion策略成功率急剧下降,且测量所得 $\alpha$ 能在零次Rollout下以0.03误差预测新噪声水平下的任务成功率。所有核心测试均预先注册,验证了基于本体锚点的性能评估指标在仿真环境中的可靠性。

关键看点 ​

  • 噪声折扣公式:策略融合本体锚点与视觉读数,权重由 $\alpha = V/(V + s^2)$ 决定,其中V为锚点误差方差且远小于线性模型预期
  • 预测精度对比:基于实测V值的模型在跨任务、学习器及噪声水平测试中MAE仅0.025,显著优于经典EIV回归的0.37
  • 闭环敏感性偏差:部署阶段策略对持续读数偏移的敏感度普遍高于理论 $\alpha$,15个合格策略无一例外,8个完全跟随偏移
  • 性能前置预测:Diffusion策略成功率随 $\alpha$ 降低而骤降,实测 $\alpha$ 可在无Rollout情况下以0.03误差预测新噪声级成功率

工程与产业观察 ​

该研究为具身智能感知鲁棒性提供了量化基准,证明模仿策略隐式利用了本体运动学信息补偿视觉噪声。提出的 $\alpha$ 预测律可作为算法选型与超参调优的前置指标,尤其对Diffusion等生成式策略具有关键指导意义,有助于在仿真阶段提前识别部署风险,减少真机试错成本。

英文摘要 (Abstract) ​

Imitation policies for robot manipulation are often trained on noisy object positions, and they learn to discount the noisy reading. We show how much they discount it, and why. Because demonstrators steer toward the object, the robot's own state carries information about where the object is. A trained policy combines this proprioceptive anchor with the reading and weights the reading by $\alpha = V/(V + s^2)$, where $s^2$ is the reading noise and $V$ the anchor's error variance. $V$ belongs to the learner and is far smaller than a linear model of the demonstrations implies, which is why classical errors-in-variables (EIV) regression overestimates $\alpha$. Measured once, $V$ predicted ten held-out sensitivities across two robomimic tasks, two learners and four new noise levels, with mean absolute error 0.025 against 0.37 for classical EIV, and a parameter-free bias test rules out the simpler explanation that the policies just overestimate the noise. The discount does not carry over to deployment, where the policy produces its own state: closed-loop sensitivity to a persistent reading offset exceeded $\alpha$ in all 15 policies that passed the registered gates, although the stricter registered prediction, that all would follow the offset fully, held for 8 of them. Post hoc, lower $\alpha$ went with lower success, steeply for diffusion, and a registered prospective test confirmed this for diffusion on Can at two new noise levels: the measured $\alpha$ predicted success before any rollout, within 0.03 (95% intervals $\pm 0.08$), although training noise predicts it equally well. Every test was registered before its data, including the classical benchmark that failed, and all other analyses are labelled post hoc; all results are in simulation with synthetic Gaussian noise.

作者:Christopher Malili

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 2611.17658:Imitation Policies Learn a Proprioceptive Anchor: A Predictive Law for Perception Noise and Its Closed Loop Cost ↗访问 2026-11-16