Smiling reduces masculinity: principal component analysis applied to facial images

Area of Science:

• Psychology
• Computer Vision
• Human Perception

Background:

• Facial expressions and perceived gender are crucial in social interactions.
• Previous research has explored factors influencing gender perception, but the role of physical facial properties in conjunction with expression remains less understood.

Purpose of the Study:

• To investigate how facial expression, specifically smiling, affects gender assignment to facial images.
• To determine if physical characteristics of the face contribute to the influence of expression on gender perception.

Main Methods:

• Computational analysis of facial images was performed.
• 36 observers rated the masculinity/femininity of 96 faces (48 male, 48 female) with neutral or smiling expressions.
• Principal Component Analysis (PCA) was applied to pixel luminance values to analyze facial features.

Main Results:

• Smiling significantly decreased perceived masculinity in male faces, especially among male observers.
• Smiling did not significantly affect perceived femininity in female faces.
• PCA revealed that the third principal component accounted for significant variance in both facial expression and perceived gender, suggesting a physical basis.

Conclusions:

• The perceived gender of a face is influenced by the interplay between facial expression and inherent physical facial traits.
• Smiling's impact on gender perception, particularly masculinity, is partly attributable to physical facial configurations.
• This suggests a complex relationship between facial dynamics and gender categorization.