sponge_marker_merged_video

该数据集是使用LeRobot工具创建的机器人学习数据集，专门用于机器人技术任务。数据集包含100个训练集，总帧数为78522，帧率为30fps。数据集中有两个任务，涉及so_follower机器人类型。数据集结构包括观察特征（如手眼图像和前方图像，分辨率均为480x640x3）、状态观察（6维浮点数组）、动作（6维浮点数组）以及时间戳、帧索引、集索引、索引和任务索引等元数据。数据以Parquet格式存储，没有视频文件。数据集遵循Apache 2.0许可证。This dataset is a robot learning dataset created using the LeRobot tool, specifically designed for robotics tasks. It contains 100 training episodes with a total of 78,522 frames at a frame rate of 30fps. The dataset includes two tasks involving the so_follower robot type. The structure comprises observation features (such as hand-eye images and front-view images, both with a resolution of 480x640x3), state observations (6-dimensional floating-point arrays), actions (6-dimensional floating-point arrays), and metadata like timestamps, frame indices, episode indices, indices, and task indices. The data is stored in Parquet format without video files. The dataset follows the Apache 2.0 license.