The COVID-19 pandemic has made it difficult to conduct research involving human-robot interactions (HRIs). As a result, many robot competitions have been canceled in the past year. To address this problem, we propose a cloud-based VR platform for the crowdsourcing of embodied cognition with multiple human users and a virtual robot. The feasibility of the platform is evaluated from the latency of the interactive events and the motions of avatars, based on the number of users logging in to the VR scene and the distance between the local and server environments. We also discuss the feasibility of our system when dealing with large-scale datasets in embodied HRI experiments.

COVID-19 大流行使进行涉及人机交互 (HRI) 的研究变得困难。因此，在过去的一年里，许多机器人比赛被取消。为了解决这个问题，我们提出了一个基于云的 VR 平台，用于众包多个人类用户和一个虚拟机器人的具身认知。该平台的可行性是从交互事件的延迟和虚拟化身的动作，基于登录到 VR 场景的用户数量以及本地和服务器环境之间的距离来评估的。我们还讨论了在体现 HRI 实验中处理大规模数据集时我们系统的可行性。