Dataset of paper Human2Robot: Learning Robot Actions from Paired Human-Robot Videos (AAAI 2026 Oral)
/end_position: eef 6-dof position, [x, y, z, roll, pitch, yaw] represent position and orientation, where the rotation is expressed in Euler angles (degrees) with the order XYZ.
/gripper_state: 0/1, 1 for gripper open，0 for close
/action: This attribute is not available in version v0… See the full description on the dataset page:
https://huggingface.co/datasets/dannyXSC/HumanAndRobot.
