When facial expressions dominate emotion perception in groups of virtual characters

Robin Palmberg, Christopher Peters, Adam Qureshi

Research output: Contribution to conferencePaper

1 Citation (Scopus)
92 Downloads (Pure)

Abstract

Virtual characters play a central role in populating virtual worlds, whether they act as conduits for human expressions as avatars or are automatically controlled by a machine as agents. In modern game-related scenarios, it is economical to assemble virtual characters from varying sources of appearances and motions. However, doing so may have unintended consequences with respect to how people perceive their expressions. This paper presents an initial study investigating the impact of facial expressions and full body motions from varying sources on the perception of intense positive and negative emotional expressions in small groups of virtual characters. 21 participants views a small group of three virtual characters engaged in intense animated behaviours as their face and body motions were varied between positive, neutral and negative valence expressions. While emotion perception was based on both the bodies and the faces of the characters, we found a strong impact of the valence of facial expressions on the perception of emotions in the group. We discuss these findings in relation to the combination of manually created and automatically defined motion sources, highlighting implications for the animation of virtual characters.
Original languageEnglish
Pages157-160
DOIs
Publication statusE-pub ahead of print - 5 Oct 2017
Event9th International Conference on Virtual Worlds and Games for Serious Applications (VS-Games) - Athens, Greece
Duration: 6 Sep 20178 Sep 2017

Conference

Conference9th International Conference on Virtual Worlds and Games for Serious Applications (VS-Games)
CountryGreece
CityAthens
Period6/09/178/09/17

Fingerprint Dive into the research topics of 'When facial expressions dominate emotion perception in groups of virtual characters'. Together they form a unique fingerprint.

Cite this