kejia98/pnp_box_franka

该数据集是通过本地多模态记录器导出创建的。数据集包含30个episodes，总计6980帧，帧率为15fps，使用Franka机器人收集。数据以parquet文件格式存储，包含图像（主图像和次要图像，分辨率均为240x320x3）、状态（包括末端执行器位置、姿态和夹爪状态）、动作（末端执行器位置、姿态和夹爪控制）、机器人末端执行器姿态命令帧ID、夹爪宽度、夹爪抓取状态、时间戳、帧索引、episode索引、索引和任务索引等特征。数据集主要用于机器人学任务，如模仿学习或强化学习，适用于训练和评估机器人控制模型。

This dataset was created using the local multimodal recorder exporter. It contains 30 episodes with a total of 6980 frames at 15fps, collected using a Franka robot. The data is stored in parquet format and includes features such as images (primary and secondary, both with resolution 240x320x3), state (including end-effector position, orientation, and gripper status), actions (end-effector position, orientation, and gripper control), Franka end-effector pose command frame ID, gripper width, gripper grasp status, timestamp, frame index, episode index, index, and task index. The dataset is intended for robotics tasks, such as imitation learning or reinforcement learning, suitable for training and evaluating robot control models.