The document discusses using random forests for skin detection from images. It provides an overview of the scheme, which uses improved Hue, Saturation, Luminance (IHLS) color space and transforms RGB values to IHLS features. Random forests showed the best performance compared to other models. The document references code and datasets, provides details on implementing random forests in R, and states results were incomplete due to a small training set and need for parallel computing. It concludes with time for questions.