Anthropomorphism Influences Perception of Computer-animated Characters' Actions
Thierry ChaminadeJessica HodginsMitsuo Kawato
Social Cognitive and Affective Neuroscience (2007)

Computer-animated characters are common in popular culture and have begun to be used as experimental tools in social cognitive neurosciences. Here we investigated how appearance of these characters' influences perception of their actions. Subjects were presented with different characters animated either with motion data captured from human actors or by interpolating between poses (keyframes) designed by an animator, and were asked to categorize the motion as biological or artificial. The response bias towards 'biological', derived from the Signal Detection Theory, decreases with characters' anthropomorphism, while sensitivity is only affected by the simplest rendering style, point-light displays. fMRI showed that the response bias correlates positively with activity in the mentalizing network including left temporoparietal junction and anterior cingulate cortex, and negatively with regions sustaining motor resonance. The absence of significant effect of the characters on the brain activity suggests individual differences in the neural responses to unfamiliar artificial agents. While computer-animated characters are invaluable tools to investigate the neural bases of social cognition, further research is required to better understand how factors such as anthropomorphism affect their perception, in order to optimize their appearance for entertainment, research or therapeutic purposes.

Thierry Chaminade, Jessica Hodgins, Mitsuo Kawato (2007). Anthropomorphism Influences Perception of Computer-animated Characters' Actions. Social Cognitive and Affective Neuroscience.

@article{ Chaminade:2007:Perception,
author = "Thierry Chaminade and Jessica Hodgins and Mitsuo Kawato",
title = "Anthropomorphism Influences Perception of Computer-animated Characters' Actions",
year = "2007",
month = may,
journal = "Social Cognitive and Affective Neuroscience",
doi = "10.1093/scan/nsm017",