qb1t/lekiwi-record-test
收藏资源简介:
该数据集是一个机器人学数据集,使用LeRobot框架创建。数据包含5个episodes,总计4500帧,以parquet文件格式存储,视频文件为mp4格式。数据集特征包括:动作(9维浮点数,表示机械臂的肩部平移、肩部提升、肘部弯曲、腕部弯曲、腕部旋转、夹爪位置以及x、y和theta速度)、观测状态(与动作相同的9维浮点数)、前视摄像头图像(480x640分辨率,3通道,30fps视频)和腕部摄像头图像(640x480分辨率,3通道,30fps视频)。此外,还包含时间戳、帧索引、episode索引、索引和任务索引等元数据。机器人类型为lekiwi_client,数据集总大小约为300MB(数据100MB,视频200MB),适用于机器人控制和视觉任务的研究与开发。
This is a robotics dataset created using the LeRobot framework. It contains 5 episodes totaling 4500 frames, stored in Parquet file format, with video files in MP4 format. The dataset features include: actions (9-dimensional floating-point numbers, representing the shoulder translation, shoulder lift, elbow flexion, wrist flexion, wrist rotation, gripper position, as well as x, y, and theta velocities of the robotic arm), observations (9-dimensional floating-point numbers identical to the action dimensions), front-facing camera images (480×640 resolution, 3 channels, 30fps video), and wrist camera images (640×480 resolution, 3 channels, 30fps video). Additionally, it includes metadata such as timestamps, frame indices, episode indices, indices, and task indices. The robotic platform type is lekiwi_client. The total size of the dataset is approximately 300 MB (100 MB for data and 200 MB for video files), and it is suitable for research and development in robotic control and vision tasks.




