28 October، 2022
Master thesis on “Face Detection and Recognition Using Deep Learning Algorithms”
A Master thesis was discussed in Department of Electrical Engineering / College of Engineering at University of Mosul entitled ” Face Detection and Recognition Using Deep Learning Algorithms ” submitted by (Mustafa Ismail Yousif Sulaiman ), Supervised By Assist. Prof. Dr. Mohammad Younis Thanoun. On Thursday, Oct. 27, 2022.
The research consists of four units, data set, processing Presets, detect and recognize faces through features extracted from the data. In this research, three models were proposed for the process of extracting and classifying features from the data, and two approaches were followed, the first approach includes building a lightweight model for face detection and discrimination using the convolutional neural network by using the least number of convolutional layers, and the second approach includes proposing two models previously trained on The dataset (Image-Net) are (VGG-16, Mobile-Net) using transfer learning technology with changing the volume of input data and the structure of the dense layers of classification, The three models were tested and evaluated by rating scales using three datasets (ORL, Georgia Tech Face), and a configured dataset.
The results showed that using the first approach in building the first model has a good test accuracy and loss function compared to the two proposed models for the second approach when used in small data sets.