Today, as seen in smart speakers, spoken dialogue technology is rapidly advancing to enable human-like interaction. However, current dialogue systems cannot pay attention not only to the content of speech, but also to the way of speaking and eye contact and facial expressions, while watching the facial expressions of the person with whom one is speaking. Therefore, this study participated in a Japanese competition called the "Dialogue Robot Competition" and attempted to develop a dialogue system that includes control of not only the content of speech but also the robot's facial expressions and gaze in order to realize a humanoid robot that can naturally interact with humans.
翻译:如今,正如聪明的演讲者所见,对话技术正在迅速推进,以使人性互动成为可能。 然而,目前的对话系统不仅不能关注言论的内容,而且不能关注说话和眼神接触以及面部表情的方式,同时关注与之交谈的人的面部表情。 因此,这项研究参加了日本的一场名为“对话机器人竞赛”的竞赛,并试图开发一个对话系统,不仅包括控制言论的内容,也包括控制机器人的面部表情和凝视,以便实现一个能够自然地与人类互动的人类机器人。