CLOT：用于全身类人机器人遥操作的闭环全球运动跟踪

Long-horizon whole-body humanoid teleoperation remains challenging due to accumulated global pose drift, particularly in full-sized humanoids. Recent learning-based tracking methods enable agile and coordinated motions but typically operate within the robot's local frame, neglecting global pose feedback, which leads to drift and instability during extended execution. CLOT presents a real-time whole-body humanoid teleoperation system that achieves closed-loop global motion tracking through high-frequency localization feedback. By synchronizing operator and robot poses in a closed loop, CLOT enables drift-free human-to-humanoid mimicry over long time horizons. However, directly imposing global tracking rewards in reinforcement learning presents challenges, which CLOT effectively addresses through its design.