Gueso/so101_cube_bin
收藏资源简介:
该数据集是一个机器人操作数据集,使用LeRobot框架创建,专门针对so101_mujoco机器人模型。数据集包含200个完整的情节(episodes),总计70,943帧,帧率为30fps,专注于单一任务(可能涉及立方体或箱子操作)。数据特征包括机器人的观测状态(6个关节位置:肩部平移、肩部升降、肘部弯曲、腕部弯曲、腕部旋转和夹爪位置)、两个摄像头图像(前摄像头和腕部摄像头,分辨率均为640x480,RGB格式),以及相应的动作控制(6个关节位置)。此外,还包含时间戳、帧索引、情节索引和任务索引等元数据。数据集以Parquet格式存储,视频文件以MP4格式存储,总数据大小约为100MB,视频文件大小约为200MB。该数据集适用于机器人学习、强化学习或模仿学习任务,特别是用于训练和评估机器人操作策略。
This is a robotic manipulation dataset developed using the LeRobot framework, specifically designed for the so101_mujoco robot model. The dataset consists of 200 complete episodes, totaling 70,943 frames at a frame rate of 30fps, and focuses on a single task (potentially involving cube or box manipulation). The data features include robotic observation states: 6 joint positions (shoulder translation, shoulder lift, elbow flexion, wrist flexion, wrist rotation, and gripper position), two camera images (front camera and wrist camera, both with 640×480 resolution in RGB format), as well as corresponding action controls (6 joint positions). Additionally, it contains metadata such as timestamps, frame indices, episode indices, and task indices. The dataset is stored in Parquet format, while the video files are stored in MP4 format. The total size of the dataset is approximately 100MB, and the total size of the video files is around 200MB. This dataset is applicable to robotic learning, reinforcement learning, or imitation learning tasks, especially for training and evaluating robotic manipulation strategies.




