用潜在空间预测动作后果，提前发现接触失败并中断执行。
ContactGuard: Pre-Contact Execution Monitoring with Action-Conditioned Latent World Models
- 基于动作条件的潜在世界模型，预测动作后视觉嵌入变化。
- 在真实场景中预接触失败预测准确率优于基线方法。
- 无需修改原策略，可直接部署为实时安全监控模块。
高接触密度的操作失败通常在机器人已发生接触后才被检测到，尤其在腕部摄像头设置下更为严重：尽管近距离视角有助于观察接触，但传统检测器反应滞后，可能导致推、错位、滑动或扰动物体。我们提出 ContactGuard，一种针对分块视觉运动策略的预接触执行监控系统。给定策略规划的动作块，ContactGuard 在潜在视觉空间中预测其短期后果，并在预测未来潜在表示表明可能失败时中止执行。其潜在世界模型通过无标签机器人轨迹训练，预测计划动作下的紧凑多视角视觉嵌入，避免像素级视频预测。随后使用少量带标签的预接触片段训练轻量级失败探测器。部署时，ContactGuard 在接触前锚定预测，沿策略自身动作滚动模型，并验证预测的接触后潜在表示。在多个真实世界的高接触操作任务中，ContactGuard 的失败预测精度高于直接和受扰动作的消融方法，并能作为预接触中止信号迁移至实际机器人，无需修改底层策略。
原文摘要 · Abstract (English)
Contact-rich manipulation failures are often detected only after the robot has committed to contact. This is especially limiting in wrist-camera setups: close gripper--object views help observe contact, but a poor approach may already push, miss, slip, or disturb the object before conventional detectors react. We introduce \emph{ContactGuard}, a pre-contact execution monitor for chunked visuomotor policies. Given the policy's planned action chunk, ContactGuard predicts its short-horizon consequence in latent visual space and aborts if the predicted future latent indicates likely failure. Its latent world model is trained from unlabelled robot trajectories to predict compact multi-view visual embeddings under planned actions, avoiding pixel-level video prediction. A lightweight failure probe is then trained from a small labelled set of pre-contact clips. At deployment, ContactGuard anchors prediction before an imminent contact event, rolls the model forward under the policy's own actions, and verifies the predicted post-contact latent. Across real-world contact-rich manipulation tasks, ContactGuard predicts failure more accurately than direct and corrupted-action ablations, and transfers to live robot as a pre-contact abort signal without modifying the underlying policy.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。