Age Estimation from Faces and Voices: A Review

Authors

  • Evelyne Moyse University of Liège

DOI:

https://doi.org/10.5334/pb.aq

Keywords:

age estimation, voice, face, own-age bias, own-race bias, own-gender bias

Abstract

Age estimation is a skill that we use in everyday life and that is also important in more specific settings such as police testimony. Researchers investigated age estimation either from faces or voices but a direct comparison between voices and faces was rather infrequently performed. This paper reviews the literature on age estimation from faces and voices and highlights some similarities and differences. Data from a series of experiments suggest that age estimation from faces and voices can be fairly accurate, but in some cases, a superiority of faces is observed. The age of participants and stimuli both seem to affect age estimation from faces and voices in the same way: (1) age estimation is more accurate for young stimuli than for older stimuli, (2) younger participants are more accurate than older participants irrespective of the age of stimuli, and (3) an own-age bias affects age estimation. By contrast, the influence of ethnicity and gender has been less documented. Available data support that ethnicity influence age estimation from faces but not clearly age estimation from voices. Regarding the effect of stimuli gender, results were opposite: in age estimation from faces, an advantage for male faces was reported whereas in age estimation from voices, the advantage is in favor of female voices.

Downloads

Published

2014-05-20