GENERATIVE ADVERSARIAL NETWORK BASED PORTRAIT BUILDING SYSTEM FOR CRIMINAL IDENTIFICATION

In today’s world the un-identification of a criminal is a major problem, causing which many of the cases to remain unsolved. So here we are using App based portrait building software from the photos of the criminals available with us. Extracting representative features of eyes, eyebrow, forehead, nose, ears, beard, hair styles etc. in various classes. In this Paper, a Generative adversarial network (GAN) [1] based Model is developed in which we train the model through a prepared database of criminal’s facial description