ABSTRACT Accurate quantification of skin color appearance is essential across diverse domains, yet it remains unclear how perceived facial color relates to corresponding colorimetric values. This study investigated how observers perceive and match the overall color appearance of facial images and textured skin patches, and how this appearance depends on viewing context and ethnicity. Seventy‐six Caucasian and Chinese observers participated in asymmetric color‐matching experiments in CIELAB space, with separate observer groups matching facial images or textured skin patches from a stimulus set of 80 Caucasian and 80 Chinese faces. Mean matches and color shifts from the colorimetric average (Δ L *, Δ a *, Δ b *) were analyzed as a function of stimulus type and ethnicity. Perceived appearance systematically deviated from the colorimetric average: facial matches were lighter and less chromatic, whereas skin patch matches were slightly lighter and more chromatic (Δ E * ab = 3.14 for faces; 1.66 for patches). Interestingly, stimulus type significantly affected chromatic shifts, and ethnicity effects emerged only for faces: Caucasian observers produced larger lightness increases, whereas Chinese observers produced greater reductions in chromaticity. Despite these shifts, simple linear models based on CIELAB means explained over 95% of the variance in matched colors. The results clarify the relationship between measured and perceived skin color and provide a practical, image‐based framework for quantifying facial color appearance in scientific and technological applications.
Lü et al. (2026) studied this question.