Facial morphometric differences across face databases: influence of ethnicities and sex

Abstract

The scientific need for standardized, high-quality facial stimuli has driven the creation of several face image databases in recent years. These stimuli are particularly important in facial asymmetry research. However, previous studies have reported facial anthropometric differences across a variety of ethnicities. This highlights the need to investigate whether these differences can also impact the use of face image databases, particularly in facial asymmetry research. In this study, we investigated facial asymmetry-based morphometric differences between the multi-ethnic Chicago Face Database (CFD) and the LACOP Face Database, which is composed of Brazilian subjects. We found reliable differences in facial asymmetry between the two databases, which were related to ethnic groups. Specifically, differences in eye and mouth asymmetry seem to drive these differences. The asymmetry-based morphometric differences among databases and ethnicities found in this study reinforce the necessity of creating multi-ethnic face databases.

Analyzing Facial Asymmetry in Alzheimer’s Dementia Using Image-Based Technology

“…Asians often have dense fat and fiber connections in the mid-face, reducing sagging and resulting in fewer superficial wrinkles [40]. Although our assessment method is not influenced by skin color, wrinkles, or nasal bridge height, it is important to acknowledge that racial differences have a significant impact on morphological variations related to facial asymmetry [41]. This could limit the comparability of facial asymmetry measurements across different ethnic groups, representing a limitation in our study.…”