Human Perception on Social Robot’s Face and Color Expression Using Computational Emotion Model

Researchers have explored the effects of expressing emotions using various modalities in the field of social robots. Prior studies have demonstrated that the use of color in emotional expressions can enhance user acceptance, relationship building, and communication effectiveness. This study aims to validate the effectiveness of different modalities in expressing Ekman’s six basic emotions. Specifically, four modalities were compared: face expression (F), face and LED color expression (FL), face and LED color expression with blinking (FLB), and face and eye color expression (FE). To accomplish this, we developed a small social robot prototype and used a computational emotion model to improve robot dynamics and interactivity. The findings revealed that, although the F modality effectively expressed emotions, ambiguous emotions were better perceived when color or blinking was incorporated. Emotions such as anger, sadness, disgust, surprise, and fear were better conveyed to the participants when the FL, FLB, and FE modalities were utilized. For happiness, F alone was sufficient for recognition. This study provides empirical evidence on the effectiveness of different modalities for expressing emotions in social robots and offers valuable insights gathered from participants’ feedback and reflections.