The Influence of Mixed-Gender Avatar Facial Features on Racial Perception: Insights from the VALID Avatar Library

Abstract

Using the VALID Avatar Library, we created 14 mixed-gender avatars by combining facial features from male and female avatars within seven racial categories. In an online study (n=68), the avatars were perceived as racially ambiguous, despite source avatars sharing the same race. Same-race participant analysis showed nuanced ambiguity reduction, emphasizing categorization refinement. Findings suggest a potential link between gender and race perception, influenced by limited perceptual expertise.

Tiffany D. Do

Assistant Professor of Computer Science