Carnegie Mellon Receptionist Corpus

本数据集名为‘卡内基梅隆接待员语料库’，由卡内基梅隆大学创建，旨在收集和分析接待员互动中的非言语行为。数据集包含60次互动记录，涉及美国英语和阿拉伯语母语者在多哈和匹兹堡大学建筑接待台的自然角色扮演。记录的行为包括目光方向、手势、头部动作、躯干位置和面部表情。数据集的创建过程涉及参与者招募、角色扮演和视频录制。该数据集主要应用于人机交互领域，特别是用于研究机器人接待员与用户之间的自然语言对话，以提高交互的自然性和效率。

This dataset, titled the Carnegie Mellon Receptionist Corpus, was developed by Carnegie Mellon University with the objective of collecting and analyzing nonverbal behaviors during receptionist interactions. It contains 60 interaction records involving natural role-play sessions between native speakers of American English and Arabic at reception desks in buildings located in Doha and on the University of Pittsburgh's campus. Recorded behaviors include gaze direction, gestures, head movements, torso positioning, and facial expressions. The development process of this dataset involved participant recruitment, role-play sessions, and video recording. This dataset is primarily applied in the field of Human-Computer Interaction (HCI), specifically for researching natural language dialogue between robotic receptionists and users to improve the naturalness and efficiency of interactions.