seriintan/rollout_eval_data_real_half_task1_20260526_151751

该数据集是使用LeRobot工具创建的，主要用于机器人学领域。数据集包含一个总集数为1的episode，总帧数为18975，以30fps的帧率采集。数据结构包括动作（action）、观察状态（observation.state）、前端图像观察（observation.images.front）、时间戳（timestamp）、帧索引（frame_index）、episode索引（episode_index）、索引（index）和任务索引（task_index）等特征。动作和观察状态涉及6个维度，分别对应肩部平移、肩部提升、肘部弯曲、腕部弯曲、腕部旋转和夹爪位置。前端图像观察为视频格式，分辨率为480x640，3个通道，使用av1编解码器。数据集以Apache 2.0许可证发布，数据文件格式为parquet，总数据文件大小为100MB，视频文件大小为200MB。机器人类型为so_follower，分割为训练集（train: 0:1）。数据集旨在支持机器人控制和研究任务。This dataset was created using the LeRobot tool, primarily for robotics applications. It contains one total episode with 18,975 frames collected at 30 fps. The dataset structure includes features such as action, observation.state, observation.images.front, timestamp, frame_index, episode_index, index, and task_index. The action and observation.state involve 6 dimensions corresponding to shoulder pan, shoulder lift, elbow flex, wrist flex, wrist roll, and gripper positions. The front image observation is in video format with a resolution of 480x640 and 3 channels, using the av1 codec. The dataset is released under the Apache 2.0 license, with data files in parquet format, total data file size of 100 MB, and video file size of 200 MB. The robot type is so_follower, and splits are defined as training set (train: 0:1). It is designed to support robot control and research tasks.