The Department of Computer Scienceheld a master thesis defense

The aim of the master thesis is to deal with a limited number of images, based on deep learning techniques by proposing novel techniques. The proposed technique includes building a hybrid model consisting of the Autoencoder Network and the Convolutional Neural Network (CNN). The experiments used the COEP database in addition to a database that was collected from different people in Najaf Governorate. The accuracy of the obtained results were 97.85% and 98.61%, respectively. The defense was accepted with success.