mbodiai/ABB_PandG_20250401_150119

这个数据集包含了机器人观察到的图像、指令、动作、状态信息以及奖励值等。具体来说，观察部分包含图像和字符串类型的指令与提示；动作部分包含机器人的姿态（包括x、y、z、roll、pitch和yaw六个维度）、抓取力度和夹爪力度；状态部分包含两个摄像头的前视和顶视图像、深度信息、世界中的物体和夹爪姿态；奖励部分是一个浮点数，表示某一动作的奖励；元数据部分包含集数索引和步骤索引。数据集分为训练集，共有5个示例，总大小为51961.8 KB。

The dataset includes robot observations such as images, instructions, actions, state information, and reward values. Specifically, the observation part contains images and string-type instructions and prompts; the action part includes the robots pose (six dimensions: x, y, z, roll, pitch, and yaw), grasp force, and gripper force; the state part contains front and top view images from two cameras, depth information, objects in the world, and gripper pose; the reward part is a float representing the reward for an action; the metadata part includes episode index and step index. The dataset is split into a training set with 5 examples, totaling 51,961.8 KB.