leokswang/19122025-foldclo-07_lerobot_format
收藏资源简介:
该数据集使用LeRobot框架创建,是一个机器人学数据集,专门针对Yam机器人。数据集包含10个训练 episodes,总计47525帧,帧率为30fps。数据以parquet格式存储,并包括以下特征: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;14维的机器人状态观察(浮点型);14维的动作数据(浮点型);以及时间戳、帧索引、episode索引等元数据。数据集主要用于机器人控制和强化学习任务,所有数据均用于训练集。
This dataset was created using the LeRobot framework and is a robotics dataset specifically for the Yam robot. It contains 10 training episodes, totaling 47,525 frames at a frame rate of 30 fps. The data is stored in parquet format and includes the following features: image observations from three cameras (top, left, right), each with a resolution of 360x640x3; 14-dimensional robot state observations (float32); 14-dimensional action data (float32); and metadata such as timestamp, frame index, and episode index. The dataset is intended for robotics control and reinforcement learning tasks, with all data allocated to the training set.




