mbodiai/ABB_PandG_20250402_101240

该数据集包含机器人交互相关的观测数据、动作数据、状态数据、奖励数据和元数据。观测数据由图像、指令和提示组成；动作数据包括姿态（包括x、y、z、roll、pitch、yaw六个维度）、抓取力度和夹爪力度；状态数据包含来自两个摄像头的图像、深度信息、世界中的物体信息和夹爪姿态序列；奖励数据为浮点数；元数据包括剧集索引和步索引。数据集分为训练集，共有5个示例，总大小为52943188字节。

The dataset includes robot interaction-related observation data, action data, state data, reward data, and metadata. Observation data consists of images, instructions, and prompts; action data includes pose (with six dimensions: x, y, z, roll, pitch, yaw), grasp force, and gripper force; state data contains images and depth information from two cameras, objects in the world, and a sequence of gripper poses; reward data is a float; metadata includes episode index and step index. The dataset is split into a training set with 5 examples, totaling 52943188 bytes.