Prototypes of facial attributes developed through image averaging techniques

S. S. Hawkins*, D. I. Perrett, D. M. Burt, D. A. Rowland, R. I. Murahata

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

2 Citations (Scopus)


Image capture and quantification has proven useful in a variety of scientific applications, for example, biology, medicine, geology, meteorology and forensics. The objective of this research was to utilize this technology to quantify clinical- and consumer-perceivable changes in facial attributes. A panel of expert assessors was trained, and, in a large consumer study, consumer facial attributes were identified and grading scales for each attribute were established. These experts then rated over 240 subjects on a total of 19 different facial attributes. Based on methodology developed by Perrett et al., facial averages or prototypes were computed from panelists rated high or low for each attribute. Prototypes were developed in a 3 step process: 1) selection of 224 predefined feature points; 2) calculation of average face shape; and 3) 'morphing' individual faces into that shape and blending the images together. Naive assessors could readily appreciate the differences in facial appearance of the prototypes. In addition, expert graders were able to identify the general class of attribute affected. This method provides a powerful tool for assessing the effects of skin care technologies.

Original languageEnglish
Pages (from-to)159-165
Number of pages7
JournalInternational Journal of Cosmetic Science
Issue number3
Publication statusPublished - 1999


  • Caricaturing
  • Facial averaging
  • Healthy glow


Dive into the research topics of 'Prototypes of facial attributes developed through image averaging techniques'. Together they form a unique fingerprint.

Cite this