ankile/real01c-marker-insert-d1-bc-r0

该数据集是一个用于机器人学习的开源数据集，基于LeRobot框架创建。数据集包含200个训练集，总计38680帧数据，使用Franka机器人收集。数据以parquet格式存储，并附带视频文件（MP4格式，15fps，分辨率480x640）。数据集特征丰富，包括机器人观察状态（如笛卡尔位置、关节位置、速度、夹爪位置）、动作（如笛卡尔速度、关节速度、夹爪动作）、奖励、成功标志、步骤计数等。此外，还提供来自两个相机视角的视频观察数据（标识为18650758_left和25916956_left）。数据集适用于机器人控制、强化学习和模仿学习等任务。This dataset is an open-source dataset for robotics learning, created using the LeRobot framework. It contains 200 training episodes with a total of 38,680 frames, collected using a Franka robot. The data is stored in parquet format and includes accompanying video files (MP4 format, 15fps, 480x640 resolution). The dataset features a wide range of attributes, such as robot observation states (e.g., Cartesian position, joint position, velocity, gripper position), actions (e.g., Cartesian velocity, joint velocity, gripper action), rewards, success flags, step counts, and more. Additionally, it provides video observations from two camera perspectives (identified as 18650758_left and 25916956_left). The dataset is suitable for tasks such as robot control, reinforcement learning, and imitation learning.