A graduate student from the Department of Computer Engineering at the University of Technology (Noor Dheyaa Jaafar) has been awarded a Master of Science degree in “Human Multiview Face Recognition System”, on Thursday, 21 February, 2019 at the conference hall of the Computer Engineering Department.

In her thesis, the researcher proposed a system capable of working in two modes (online, and offline), for recognizing the person. The multiple images from the FEI database are used for performance evaluation of the face detection and recognition process in the offline mode. Three cameras have been connected in the online mode by using Raspberry Pi 3 platform with the purpose of improving processing time.

The front face detection has been done using Viola and Jones algorithm, while the facial side images are being detected using the skin detection methodology. Histogram oriented gradiant (HOG) algorithm is employed for extracting features of an image. For the recognition step, the image is passed through two algorithms, i.e., once by principal component analysis (PCA) algorithm, and again by support vector machine (SVM), in order to compare the performance in both cases.

Deferent pre-processing techniques have been used, such as image cropping and image resizing. Each technique applied to the face image to improve the quality leads to increase the recognition rates.

In the online mode, the highest accuracy rate obtained was 92.12% in SVM method when the image size is (128×128) pixels. Moreover, 87% accuracy is achieved in the PCA recognition method. While in offline mode, the highest accuracy rate obtained was 95% with SVM method when the image size is (64×64) pixels. Moreover, 90% of accuracy is achieved by the PCA recognition method.

The discussion committee consisted of five members, Asst. Prof. Dr. Bassim AbdulBaki Jumaa from the University of Technology – Iraq / Computer Engineering Department as Chairman, Asst. Prof. Dr. Buthaina Fahran Abdel from the University of Information Technology / Smart Cities Department as a member, Asst. Prof. Dr. Ahmed Ibrahim Abdelkarim from the University of Technology – Iraq / Department of Control and Systems Engineering as a member, Dr. Saman Hameed Amin as a member and supervisor, and Dr. Ayman Dawood Salman as a member and supervisor from the University of Technology – Iraq / Computer Engineering Department.

A graduate student from the Department of Computer Engineering at the University (Duha Hassan) qualified for and was admitted to the award of Master degree on 31-01-2019 at the conference hall of the Computer Engineering Department. The title of the thesis was: “the design and implementation of the object detection system using deep learning”

In this thesis, the researcher addressed the design of a system that provides an effective model with a simplified structure that uses less memory and processing power while maintaining good accuracy and speed. This system allows the computer to tell us directly that this image contains a certain object and where it is located inside the image. The proposed system depends on the extraction of the features of the object using (CNN) and then work on these characteristics or attributes to predict the dimensions that represent the location of the object and predict the type of object between the tested species and the use of another network containing more than one exit.

The system has been able to detect highly variable objects, lighting conditions and trend. The model (48.61 MAP) has provided a detailed discussion and insight into future work, which can be interesting for both academia and industry, especially in the field of learning and computer vision.

The discussion committee consisted of Prof. Dr. Emad H. Al-Husseini from the University of Technology / Computer Engineering Department as Chairman, Asst. Prof. Dr. Amjad J. Hamidi from the University of Technology / Department of Control and Systems Engineering as member, Asst. Prof. Dr. Muthana H. Hamad from Al- Mustansiriya University / College of Engineering / Computer Engineering Department as a Member, Asst. Prof. Dr. Bassem A. Jumaa and Dr. Saman H. Amin from the University of Technology / Computer Engineering Department as Members and Supervisors.