Abstract: A head image from a source image based on a model of the source image as a mixture of at least two image layers. One or more of the image layers are components of a foreground corresponding to the source head image and one or more other ones of the image layers are components of a background corresponding to parts of the source image outside the source head image. The source head image is color-adjusted based on skin regions of a target head image in a target image. The color-adjusted source head image is geometrically transformed to conform to the target head image in pose and size. A composite image is generated. In this process, the geometrically transformed and color-adjusted source head image is composited into a version of the target image at a location corresponding to the target head image.