Interactive Evolutionary Strategies for Automatic Facial Composite Generation

The main aim of this project is to develop and anal yze a program for the automatic face generation through i nteractive evolutionary strategies (IES). The user selects a n umber of images based on their similarity to a target image from a set of images presented to him via the interface. The gene ration of the facial images at each iteration are achieved by per forming parent selection, recombination, mutation and surv ivo selection steps of evolutionary strategies (ES) and optimizing the T dimensional parameter vector of an active app e rance model based on the selections of the user interacti vely. The main goal is to continue the iterations until the u ser is satisfied with at least one of the generated images. The prog ram has been implemented and the preliminary tests show that IES gives successful results.