CoRL2026-CSI/SO101-cap_close_drawer_10fps_50epi_8to2

该数据集是一个机器人控制数据集，专门用于机器人学习和模仿任务。数据集包含50个训练集片段，总计11712帧数据，帧率为10fps。数据集中涉及一个任务，即cap_close_drawer（关闭抽屉）。数据集的结构包括多种特征，如机器人的状态观测（6维关节位置）、动作（6维关节控制）、来自顶部和左腕摄像头的图像观测（分辨率480x640，RGB三通道）、末端执行器位置（6维XYZRPY坐标）、二进制夹持器状态、技能信息（自然语言描述、验证问题、类型、进度、目标位置）、子任务信息（自然语言描述、对象名称、目标位置）以及时间戳、帧索引、片段索引等元数据。数据集以Parquet格式存储，总数据文件大小为100MB，视频文件大小为200MB。该数据集适用于机器人控制、模仿学习和强化学习等研究领域。This dataset is a robotic control dataset specifically designed for robotic learning and imitation tasks. It contains 50 training episodes, totaling 11712 frames at a frame rate of 10fps. The single task involved is cap_close_drawer (closing a drawer). The dataset structure includes various features: robot state observations (6-dimensional joint positions), actions (6-dimensional joint control commands), image observations from the top and left-wrist cameras (resolution 480x640, 3-channel RGB), end-effector positions (6-dimensional XYZRPY coordinates), binary gripper state, skill information (natural language descriptions, validation questions, type, progress, target positions), subtask information (natural language descriptions, object names, target positions), as well as metadata such as timestamps, frame indices, and episode indices. The dataset is stored in Parquet format, with a total data file size of 100MB and a video file size of 200MB. This dataset is applicable to research fields such as robotic control, imitation learning, and reinforcement learning.