Quantifying influence from form manipulation of artificial facial expression to viewers

Shyue Ran Li, Yi Chen Chen, Kuohsiang Chen, Chun Heng Ho

Research output: Contribution to journalArticlepeer-review

Abstract

Creating facial expressions manually needs to be enhanced through a set of easy operating rules, which involves adjectives and the manipulating combinations of facial design elements. This article tries to analyse viewers' cognition of artificial facial expressions in an objective and scientific way. We chose four adjectives – ‘satisfied’, ‘sarcastic’, ‘disdainful’, and ‘nervous’ – as the experimental subjects. The manipulative key factors of facial expressions (eyebrows, eyes, pupils, mouth and head rotation) were used to create permutations and combinations and to make 81 stimuli of different facial expressions with a 3-D face model in order to conduct a survey. Next, we used Quantification Theory Type I to find the best combinations that participants agreed on as representing these adjectives. The conclusions of this research are that: (1) there are differences in adopting facial features between creating artificial characters' expressions and recognising real humans' expressions; (2) using survey and statistics can scientifically analyse viewers' cognition of facial expressions with form changing; and (3) the results of this research can promote designers' efficiency in working with subtler facial expressions.

Original languageEnglish
Pages (from-to)313-329
Number of pages17
JournalDigital Creativity
Volume25
Issue number4
DOIs
Publication statusPublished - 2014 Oct 2

All Science Journal Classification (ASJC) codes

  • Arts and Humanities (miscellaneous)
  • Human-Computer Interaction
  • Computer Graphics and Computer-Aided Design
  • Computational Theory and Mathematics

Fingerprint Dive into the research topics of 'Quantifying influence from form manipulation of artificial facial expression to viewers'. Together they form a unique fingerprint.

Cite this