2D mikroanimacije obraza in analiza prepoznave čustev

Stres, Sinja (author), Gabrijelčič Tomc, Helena (mentor)

Čustva so v vsakdanjem življenju zelo pomemben del komunikacije, zato je pomembno, da jih znamo prepoznati. Izražajo namreč, kako se naš sogovornik počuti v določenem trenutku, če je vesel, žalosten, jezen, presenečen, ali pa se je morda prestrašil. Tudi ko govorimo o animiranih risankah, oglasih, objavah ipd., je zelo pomembno, da je zaznava prikazanega čustva zelo dobra. Z dobro zaznavo prikazanih čustev v animiranih vsebinah bomo namreč dosegli veliko boljšo vživetost in povezanost med likom in gledalcem, hkrati pa bo naše sporočilo bolje sprejeto in bolj zapomljivo. Z nekateri prikazi čustev v animacijah velikokrat pretiravajo, medtem ko nekateri drugi ne pokažejo dovolj intenzivnosti prikaza čustva, da bi se lahko v animacijo gledalec vživel. V raziskavi je bila analizirala zaznava animirane vizualizacije čustev na obrazu ilustriranega lika. Primerjala se je zaznava šestih osnovnih prikazov čustev, in sicer prikaz veselja, žalosti, gnusa, jeze, presenečenja in strahu. Vsako čustvo se je preko tehnik mikroanimiranja prikazalo v treh intenzivnostih, in sicer v zelo blagi intenzivnosti, v srednji intenzivnosti in v pretirani intenzivnosti prikazanega čustva. Zaznava posameznih čustev je bila analizirana s pomočjo ankete, kjer se je poleg pravilnosti odgovorov beležila tudi hitrost odgovora in jasnost tega. Na izbranih rezultatih so bile analizirane razlike v zaznavi posameznih čustev pri določenih intenzitetah. V rezultatih je predstavljeno, kako se je zaznava razlikovala glede na izbrano čustvo in intenziteto prikaza tega. Iz tega je bilo poglobljeno razumevanje o učinkih mikroanimacij in definirane so bile smernice, ki so v pomoč animatorjem pri vprašanju, katera čustva je treba še posebej poudariti, da so dobro zaznana in katera so lahko precej subtilna, vendar je zaznava še vedno zelo visoka.

• Language: Slovenian
• Keywords: 2D animacija, mikroaniamcija, prikazovanje čustev, mikroizrazi, makroizrazi.
• Work type: Master's thesis/paper
• Organization: NTF - Faculty of Natural Sciences and Engineering
• Year: 2022
• PID: 20.500.12556/RUL-139791
• Publication date in RUL: 07.09.2022

Secondary language

• Language: English
• Title: 2D facial microanimation and analysis of emotion perception

Abstract

Emotions represent a big part of everyday communication, that’s why it’s important for us to learn how to identify them. They communicate how a person feels, if they’re happy, sad, angry, surprised or even scared. Even when talking about animated cartoons, advertisements, posts, etc., it’s important that we keep the emotion recognition high. That ensures a better connection between the character and the viewer, as well as making our message more memorable. Some facial expressions depicting emotions are often exaggerated, while others aren’t emphasised enough to ensure good emotional recognition. This study analyzes the perception of animated facial expressions depicting different emotions on an illustrated character. We compared the perception of six basic emotional expressions (happiness, sadness, anger, surprise, fear and disgust). Using micro-animation techniques, each emotion was depicted in three intensity levels (a subtle version, a normal version and an exaggerated version). Emotion perception was analysed with a survey where speed and accuracy were measured in addition to the correctness of the answers. The results show how the perception changed based on the emotion and the intensity shown and grant us a deeper understanding of micro-animations and facial expression. Based on the results, guidelines were formed to help animators answer the question, which emotions need to be especially exaggerated to be perceived and which emotions can be animated more subtly while still maintaining good emotional perception.

• Keywords: 2D animation, micro-animation, displaying emotions, micro expressions, macro expressions.