HRI

该数据集包含了在研究人机交互过程中，为了探究困惑检测而收集的Pepper机器人音频记录、面部视频和姿态视频。此外，数据集还包括了用于估算情绪、眼动和头部姿态的面部框架数据，并附带了一份研究后使用的五级李克特量表的问卷调查。该数据集的规模涉及29名参与者，其任务聚焦于在人机交互中的困惑检测。

This dataset contains audio recordings, facial videos, and pose videos of the Pepper robot collected during human-robot interaction (HRI) studies investigating confusion detection. Additionally, it includes facial framework data for estimating emotions, eye movements, and head poses, as well as a post-study questionnaire employing a 5-point Likert scale. A total of 29 participants were involved, with the study tasks focused on confusion detection during human-robot interaction.