leokswang/22122025-foldclo-12_lerobot_format
收藏资源简介:
该数据集使用LeRobot创建,是一个机器人学数据集,专门针对yam机器人类型。它包含10个episodes,总计30417帧,帧率为30fps,所有数据用于训练。数据集以parquet格式存储,每个episode分块组织。特征包括来自三个摄像头(顶部、左侧、右侧)的图像观察,每个图像分辨率为360x640x3(高度、宽度、通道),以及14维浮点数组的状态观察和动作数据。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。数据集结构支持机器人学习任务,如动作预测或状态估计,但具体任务未在README中明确说明。
This dataset was created using LeRobot and is a robotics dataset specifically for the yam robot type. It contains 10 episodes, totaling 30,417 frames at 30 fps, with all data used for training. The dataset is stored in parquet format, with each episode organized into chunks. Features include image observations from three cameras (top, left, right), each with a resolution of 360x640x3 (height, width, channels), as well as state observations and action data as 14-dimensional float32 arrays. Additionally, metadata such as timestamp, frame index, episode index, index, and task index are included. The dataset structure supports robotics learning tasks, such as action prediction or state estimation, though specific tasks are not explicitly detailed in the README.



