ALOHA demonstration data

该数据集是一个高度多模态、带有语言标注的数据集，专门用于训练机器人在灵巧操作任务上的策略。它包含了针对各项任务的语言指令，旨在提高机器人策略的性能。该数据集的规模涉及多个任务和不同的机器人形态，具体任务包括拾放、笔盖开启、寿司切割、烤面包以及擦拭等多项机器人操作任务。

This is a highly multimodal, language-annotated dataset specifically developed for training robot policies on dexterous manipulation tasks. It includes task-specific language instructions to enhance the performance of robot policies. This dataset encompasses multiple tasks and diverse robot morphologies, with specific robotic manipulation tasks including pick-and-place, pen cap removal, sushi cutting, toast preparation, wiping, and numerous other relevant tasks.