Human-in-Loop Interactive Social Robot Navigation Environment

该数据集为模拟环境，旨在评估社交机器人使用人类语言命令和反馈进行导航性能的表现。该数据集涵盖了不同交互式任务类型，包括人引导、人跟随以及点对点导航，为这些任务的评估提供了便利。规模上，该数据集包含了1万个训练集以及每种任务500个随机案例。任务内容是在复杂环境中，结合人际互动的社交机器人导航。

This dataset is a simulated environment designed to evaluate the navigation performance of social robots when utilizing human language commands and feedback. It covers various interactive task types, including human-guided navigation, human-following navigation, and point-to-point navigation, which facilitates the evaluation of these tasks. In terms of scale, the dataset includes 10,000 training samples and 500 randomly generated cases for each individual task. The core of these tasks is social robot navigation in complex environments with interpersonal interactions.