Skip to Main content Skip to Navigation
Journal articles

Emotional Valence Recognition on Virtual, Robotic, and Human Faces: a Comparative Study

Abstract : With the advent of new technologies in everyday life, leisure, or therapy, we will increasingly interact with a non-human virtual character. Understanding facial expressions and intentions of these virtual agents is important to enable them to achieve their goals. The objective of our study is to assess whether expressions are perceived as being positive or negative on faces more or less similar to those of humans. Eighty-three undergraduate students took part in a computerized emotion recognition task. The participants had to identify whether each face expressed a positive or a negative emotion. Eight different faces (human, avatar, mesh, and robot) were shown 38 times each on a computer screen. Each face was represented by a photo. Response time and the number of correct responses were recorded. Our research has raised important points: the accuracy and time taken for emotion recognition were found to be similar on human or avatar faces. On the other hand, as soon as these faces were too ambiguous or schematic, emotion recognition capacities were found to be diminished.
Document type :
Journal articles
Complete list of metadata
Contributor : Lisa Cerda Connect in order to contact the contributor
Submitted on : Thursday, October 8, 2020 - 8:38:22 AM
Last modification on : Wednesday, November 3, 2021 - 7:30:45 AM

Links full text




Lisa Cerda, Pierluigi Graziani, Jonathan Del-Monte. Emotional Valence Recognition on Virtual, Robotic, and Human Faces: a Comparative Study. Journal of Technology in Behavioral Science, Springer, 2020, ⟨10.1007/s41347-020-00172-5⟩. ⟨hal-02960854⟩



Record views