mimicgen_mugclenup_d0_224x224

该数据集使用LeRobot框架创建，包含Panda机器人在模拟环境（robosuite）中执行杯子清理任务（mimicgen_mugclenup）的演示数据。数据集包含978个episodes，总计330404帧，每个episode包含动作（7维，包括位置增量、旋转增量、夹爪控制）、观测状态（如末端执行器位姿、关节状态、物体状态等）以及两个224x224的RGB图像（agentview和eye_in_hand）。数据用于机器人模仿学习，采用CC-BY-4.0许可。This dataset is created using the LeRobot framework, containing demonstration data of a Panda robot performing a mug cleaning task (mimicgen_mugclenup) in a simulated environment (robosuite). The dataset includes 978 episodes totaling 330,404 frames. Each episode contains actions (7-dimensional, including position delta, rotation delta, gripper control), observations (e.g., end-effector pose, joint states, object states), and two 224x224 RGB images (agentview and eye_in_hand). The data is intended for robot imitation learning and is licensed under CC-BY-4.0.