Analysis of emotion recognition through 2D micro-animations of an illustrated character's face

Emotions make up a large part of everyday communication. Humans learn to recognize emotions by observing others and by referencing their feelings with the emotions of other people. Also, in cartoons, commercials, posts, etc., it's important that the design of the characters keeps the recognition of emotions high. Expressed emotions provide a better connection between the character and the viewer, making the message more understandable and tangible. This study analyses the recognition of animated facial expressions depicting different emotions on the face of an illustrated character. The accuracy of rec-ognition of six basic emotional expressions (joy, sadness, anger, surprise, fear, and disgust) was compared. Using micro-animation techniques, each emotion was presented in three levels of intensity (a subtle version, a normal version, and an exaggerated version). Emotion recognition was analysed with a meth-od of metric analysis of viewing and surveying that measured recognition time and accuracy in addition to the correctness of the characters' emotion recognition. Statistically relevant differences between the results of animated emotion recognition as a function of recognition time and type of recognition task were examined. The results show how recognition changed as a function of the emotion shown and intensity, and provide a deeper understanding of micro-animations and facial expressions on the animated character's face. Statistically relevant differences were found especially in the recognition of the emotions disgust and anger compared to the recognition of the emotions joy, surprise, fear. Based on the results, guidelines are given to help animators answer the question of which emotions need to be particularly exaggerated to be correctly recognised and which emotions can be animated more subtly without affecting emotional perception.