Rajen and Abhinav (2012) addressed the challenge of detecting skin-like
regions in images as a component of the intricate process of facial recognition.
To achieve this goal, they curated the “Skin segmentation” data set, comprising
RGB (R-red, G-green, B-blue) values of randomly selected pixels from
N = 245,057 facial images, including 50,859 skin samples and 194,198 nonskin
samples, spanning diverse age groups, racial backgrounds, and genders.