sonya-1-23/Panda_EE_Fast_Cube_filtered
收藏资源简介:
该数据集是使用LeRobot创建的机器人学数据集,专为机器人控制和模仿学习任务设计。数据集包含48个episodes,总计26008个frames,覆盖2个不同的任务。数据以parquet格式存储,总数据文件大小为100MB,视频文件大小为200MB,帧率为30fps。数据集仅包含训练集(splits: train)。数据特征包括:动作(action),为7维浮点数组,表示末端执行器的姿态增量(x、y、z、rx、ry、rz)和夹爪命令;状态观测(observation.state),为15维浮点数组,包括7个关节位置、夹爪状态归一化值、夹爪命令二进制值以及末端执行器姿态(位置和旋转);图像观测(observation.images.camera_1和camera_2),均为视频格式,分辨率为480x640,3通道(RGB),使用av1编解码器;此外还包括时间戳、帧索引、episode索引、索引和任务索引等元数据。数据集基于Apache 2.0许可证发布,适用于机器人模拟和真实世界控制研究。
This dataset is a robotics dataset created using LeRobot, designed for robot control and imitation learning tasks. It contains 48 episodes with a total of 26008 frames, covering 2 distinct tasks. The data is stored in parquet format, with a total data file size of 100MB and video file size of 200MB, at a frame rate of 30fps. The dataset includes only a training set (splits: train). Features include: action, a 7-dimensional float32 array representing delta end-effector pose (x, y, z, rx, ry, rz) and gripper command; observation.state, a 15-dimensional float32 array with 7 joint positions, normalized gripper state, binary gripper command, and end-effector pose (position and rotation); image observations (observation.images.camera_1 and camera_2), both in video format with a resolution of 480x640, 3 channels (RGB), using the av1 codec; additionally, metadata such as timestamp, frame index, episode index, index, and task index are included. The dataset is released under the Apache 2.0 license and is suitable for robotics simulation and real-world control research.




