mbodiai/ABB_PandG_20250331_212231

该数据集包含机器人执行任务时的观察、动作、状态和奖励等信息。观察包括图像和指令，动作包括姿态（包括x、y、z、roll、pitch、yaw六个维度）、抓取力度和夹爪力度。状态包括来自两个摄像头的图像、深度信息、世界中的物体和夹爪姿态。此外，每个步骤都有一个奖励值和元数据信息，如集数索引和步骤索引。数据集分为训练集，共有6个示例。

The dataset contains information about observations, actions, states, and rewards during robot execution of tasks. Observations include images and instructions, actions include pose (with six dimensions: x, y, z, roll, pitch, yaw), grasp force, and gripper force. States include images from two cameras, depth information, objects in the world, and gripper pose. Additionally, each step has a reward value and metadata such as episode index and step index. The dataset is split into a training set with a total of 6 examples.