HyeonseokE/SO101-DualArm-teleop_fold_towel_40epi
收藏资源简介:
该数据集是一个用于机器人学习的数据集,专注于双机械臂折叠毛巾的任务。它使用LeRobot框架创建,包含40个训练集,总计26114帧,帧率为30fps。数据集结构包括动作和观察特征:动作特征由12个浮点值组成,表示左右机械臂的关节位置(如肩部平移、肩部升降、肘部弯曲、腕部弯曲、腕部滚动和夹爪位置);观察特征包括状态观察(同样为12个浮点值的关节位置)和图像观察,图像来自三个摄像头(左腕、左顶部、右腕),分辨率为480x640,3通道,视频编码为h264。此外,数据集还包含时间戳、帧索引、集索引等元数据。数据文件以Parquet格式存储,视频文件以MP4格式存储,总数据文件大小为100MB,视频文件大小为200MB。数据集适用于机器人控制、模仿学习或强化学习等任务。
This dataset is designed for robotics learning, specifically focusing on a dual-arm towel folding task. It was created using the LeRobot framework and consists of 40 training episodes, totaling 26114 frames at a frame rate of 30fps. The dataset structure includes action and observation features: the action feature comprises 12 float32 values representing joint positions for both left and right robotic arms (e.g., shoulder pan, shoulder lift, elbow flex, wrist flex, wrist roll, and gripper positions); the observation features include state observations (also 12 float32 values for joint positions) and image observations from three cameras (left wrist, left top, right wrist) with a resolution of 480x640, 3 channels, and h264 video encoding. Additionally, the dataset contains metadata such as timestamps, frame indices, and episode indices. Data files are stored in Parquet format, while video files are in MP4 format, with total data file size of 100MB and video file size of 200MB. This dataset is suitable for tasks like robot control, imitation learning, or reinforcement learning.




