Automatic classification of human facial features based on their appearance.

Area of Science:

• Computer Vision
• Anthropometry
• Human-Computer Interaction

Background:

• Existing human body part classification systems are abundant, but facial feature taxonomies are scarce.
• Classifying facial features is challenging for human observers, leading to low inter- and intra-observer agreement.
• Applications in ergonomics, forensic anthropology, and human-machine interaction necessitate robust facial feature classification.

Purpose of the Study:

• To present a computer-based procedure for automatic facial feature classification.
• To address the difficulties and inconsistencies associated with human judgment in facial feature categorization.
• To facilitate the development of comprehensive facial feature taxonomies.

Main Methods:

• A computer-based procedure was developed to classify facial features based on their global appearance.
• The method aims to provide objective and consistent classification, overcoming human observer variability.
• The procedure was applied to generate taxonomies for eyes, mouths, and noses.

Main Results:

• The computer-based procedure successfully classified facial features automatically.
• The developed method demonstrated efficacy in generating taxonomies for specific facial features.
• This approach offers a reliable alternative to subjective human evaluations for facial feature classification.

Conclusions:

• The proposed computer-based procedure effectively classifies facial features, enhancing objectivity and consistency.
• This method facilitates the creation of valuable facial feature taxonomies for diverse scientific and technological fields.
• The automatic classification system provides a foundation for improved human-machine interaction and other applications.