fecasado/burger-to-plate-good

该数据集是一个机器人控制数据集，由LeRobot创建，用于机器人学习任务。数据集包含70个episodes，总计105962帧数据，以15fps的帧率记录。特征包括动作（action）和观测（observation），其中动作维度为26，涵盖左臂和右臂的线性与角速度、手部关节控制；观测包括状态（55维，涉及双臂和手部关节位置与力、视线数据）以及多视角图像视频（左、右、用户、用户视线视角，分辨率480x640，3通道）。数据集使用Apache 2.0许可证，机器人类型为blueberry_ros，数据以parquet和mp4格式存储，适用于训练机器人控制模型。This dataset is a robotics control dataset created by LeRobot for robot learning tasks. It contains 70 episodes with a total of 105,962 frames recorded at 15 fps. Features include actions (26 dimensions, covering linear and angular velocities of left and right arms, and hand joint controls) and observations (state with 55 dimensions, involving positions and efforts of arm and hand joints, gaze data; and multi-view image videos from left, right, user, and user gaze perspectives, with resolution 480x640 and 3 channels). The dataset is licensed under Apache 2.0, uses the blueberry_ros robot type, and is stored in parquet and mp4 formats, suitable for training robot control models.