Meta-world benchmark tasks

该数据集包含了在Meta-world环境中用于验证TemporalOT方法与其他基线方法有效性的一系列不同任务。这些任务中，每个episode的目标位置都会发生变化，且仅为强化学习智能体提供了两个专家视频演示。该数据集规模涵盖九个独特的任务，旨在通过专家视频演示来实现机器人策略学习。

This dataset comprises a series of distinct tasks within the Meta-world environment, which are used to validate the effectiveness of the TemporalOT method and other baseline approaches. For each of these tasks, the target position varies across different episodes, and only two expert video demonstrations are provided for the reinforcement learning agent. This dataset covers nine unique tasks overall, aiming to facilitate robotic policy learning through expert video demonstrations.