论文HF Daily Papers · arXiv 2610.373936 天前

QCP策略实现机器人质量属性推理时可控与迭代优化

提出Quality-Controllable Policy（QCP），支持推理时动态调节执行速度与平滑度等质量属性。通过迭代事后重标记机制扩展可控范围，在Robomimic及真实传送带、酒杯搬运任务中验证了吞吐量与成功率的双重提升。

AI 整理

仿真与训练
具身模型

本文由 AI 根据公开内容整理，可能有误，请以原文为准：Learning Quality-Controllable Robot Policies with Iterative Hindsight Relabeling（HF Daily Papers · arXiv 2610.37393）

核心内容

针对机器人策略评估仅依赖二元成功信号而忽视执行质量的问题，研究提出Quality-Controllable Policy（QCP）。该方法允许在推理阶段对速度、平滑度等单一或组合质量属性进行动态控制。为解决初始演示数据限制可控范围的问题，引入迭代事后重标记循环：收集目标标签下的在线轨迹，将成功样本加入数据集并根据实测质量重标记，进而微调策略。实验表明QCP在Robomimic和RoboEval基准上持续提升任务质量与成功率，生成轨迹质量超越初始演示。

在物理部署验证中，QCP展现了明确的工程价值。在传送带操作场景下，通过调节速度属性有效提升了系统吞吐量；在酒杯搬运任务中，控制运输稳定性显著提高了操作成功率。技术方案核心在于将抽象的质量指标转化为可学习的条件输入，并通过在线数据闭环不断修正策略对质量属性的响应边界，使机器人能够根据实际工况需求灵活权衡执行效率与稳定性，解决了传统模仿学习策略僵化、难以适应多变现场约束的痛点。

关键看点

推理时质量控制：QCP支持在推理阶段动态调节速度、平滑度等单一或组合质量属性，突破传统策略固定行为模式限制
迭代事后重标记：通过在线收集轨迹并按实测质量重标记微调，使策略生成的成功轨迹质量持续超越初始演示数据上限
真实场景效能验证：在传送带操作中调控速度提升吞吐量，在酒杯搬运中调控稳定性提高成功率，证实质量可控具实用价值
基准测试表现：在Robomimic与RoboEval实验中，QCP随迭代次数增加同步提升任务执行质量指标与整体成功率

工程与产业观察

QCP将执行质量从离线评价指标转化为在线可控变量，为具身智能在非结构化环境中的自适应部署提供新范式。其迭代重标记机制解决了高质量演示数据稀缺难题，对工业装配、精密搬运等需兼顾效率与安全性的场景具有直接参考价值，推动模仿学习从“能做”向“做好”演进。

英文摘要 (Abstract)

Robot policy evaluation often relies on a binary success signal for task completion. In real-world deployment, however, desired execution behaviors are also defined by quality attributes such as execution speed and motion smoothness, which need to be dynamically controlled according to task requirements. To address this, we introduce Quality-Controllable Policy (QCP), a robot policy that enables inference-time control over both individual quality attributes and their combinations. To extend the controllable quality range beyond initial demonstrations, we further propose an iterative hindsight relabeling loop: QCP collects online rollouts under target quality labels, adds successful trajectories to the dataset, relabels the accumulated trajectories in hindsight according to their measured quality, and fine-tunes the policy. Experiments on Robomimic and RoboEval demonstrate that QCP consistently improves both task execution quality and success rates across iterations, producing successful rollouts that exceed the quality of the initial demonstrations. Finally, we validate QCP on real-world manipulation tasks, where controlling speed increases throughput in conveyor operations and controlling transport stability improves success when carrying wine glasses, demonstrating that quality-controllable policies yield practical gains in physical deployment.

Hugging Face Daily Papers 热度：6 upvotes 作者：Sinjae (Cason) Kang, Dong Kyu Shin, Jaehong Kim, Kamin Lee, Hyoeun Kim, Kaixin Wang, Li Zhao, Seonghyeon Ye, Kimin Lee