通过反事实失败合成学习可操作的操纵恢复
Learning Actionable Manipulation Recovery via Counterfactual Failure Synthesis
AI总结：
本文提出Dream2Fix框架，通过生成反事实失败场景数据提升机器人故障恢复能力，实现高精度的故障诊断与轨迹修正。
摘要：
尽管近期基础模型显著提升了机器人操作能力，但这些系统仍难以自主恢复执行错误。现有失败学习方法依赖于昂贵且不安全的真实世界数据收集或基于模拟器的扰动，导致严重的仿真到现实差距。此外，现有视觉分析器主要输出粗略的二元诊断，而非实际恢复所需的可执行轨迹级修正。为弥合故障诊断与可操作恢复之间的差距，我们引入Dream2Fix框架，该框架直接从成功的真实世界演示中合成逼真的反事实失败场景。通过在生成世界模型中扰动动作，Dream2Fix创建配对的失败-修正数据，而无需依赖模拟器。为确保生成的数据对机器人学习具有物理可行性，我们实施了一种结构化的验证机制，严格过滤任务有效性、视觉一致性和运动学安全性。该引擎生成超过12万对配对样本的高保真数据集。使用此数据集，我们微调视觉-语言模型，联合预测故障类型和精确的恢复轨迹，将视觉异常直接映射到纠正动作。广泛的现实世界机器人实验表明，我们的方法在纠正精度上达到当前最佳水平，从19.7%提升到81.3%，并在物理部署中实现了零样本闭环失败恢复。
英文摘要：
While recent foundation models have significantly advanced robotic manipulation, these systems still struggle to autonomously recover from execution errors. Current failure-learning paradigms rely on either costly and unsafe real-world data collection or simulator-based perturbations, which introduce a severe sim-to-real gap. Furthermore, existing visual analyzers predominantly output coarse, binary diagnoses rather than the executable, trajectory-level corrections required for actual recovery. To bridge the gap between failure diagnosis and actionable recovery, we introduce Dream2Fix, a framework that synthesizes photorealistic, counterfactual failure rollouts directly from successful real-world demonstrations. By perturbing actions within a generative world model, Dream2Fix creates paired failure-correction data without relying on simulators. To ensure the generated data is physically viable for robot learning, we implement a structured verification mechanism that strictly filters rollouts for task validity, visual coherence, and kinematic safety. This engine produces a high-fidelity dataset of over 120k paired samples. Using this dataset, we fine-tune a vision-language model to jointly predict failure types and precise recovery trajectories, mapping visual anomalies directly to corrective actions. Extensive real-world robotic experiments show our approach achieves state-of-the-art correction accuracy, improving from 19.7% to 81.3% over prior baselines, and successfully enables zero-shot closed-loop failure recovery in physical deployments.
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。
