oxe-auge/fractal20220817_data_train_65000_67500_augmented

fractal20220817_data_train_65000_67500_augmented是一个用于机器人学习和增强学习的大规模数据集。它包含多种机器人（images, jaco, kinova3, kuka_iiwa, panda, sawyer, ur5e, xarm7）的图像和状态信息，以及自然语言指令。数据集以3帧每秒的速度采集，共2500个episodes，107824帧。数据特征包括机器人图像（256×320×3）、关节位置、末端执行器位置、自然语言指令等。数据以parquet格式存储，并提供了视频路径。该数据集基于OXE-AugE项目，旨在支持跨机器人策略学习的研究。

The fractal20220817_data_train_65000_67500_augmented is a large-scale dataset for robotics and reinforcement learning. It contains images and state information of various robots (images, jaco, kinova3, kuka_iiwa, panda, sawyer, ur5e, xarm7), along with natural language instructions. The dataset is collected at 3 frames per second, totaling 2,500 episodes and 107,824 frames. Features include robot images (256×320×3), joint positions, end-effector poses, natural language instructions, etc. Data is stored in parquet format, with video paths provided. This dataset is part of the OXE-AugE project, designed to support research in cross-embodiment policy learning.