Assistant Professor Dr. (Shaimaa Safaa El-Din Baha El-Din), a lecturer from the Department of Computer Engineering / College of Engineering / University of Nahrain, participated in the discussion committee of the master's student (Adhraa Khaled Dawood Hamid) from the Department of Network Engineering / College of Information Engineering / University of Nahrain, which was held on Tuesday 11/2/2025 for her thesis entitled:

Optimized AI algorithm for edge computing environments

The thesis submitted by the student aims to develop an artificial intelligence algorithm called YOLOv8, which is improved for edge computing, to reduce the size of the artificial intelligence model in memory-constrained edge devices. The thesis presented the use of the pruning and deep learning method to improve memory, which made it possible to efficiently deploy convolutional neural networks on edge devices of the Internet of Things. Experiments were also conducted on the YOLOv8n model for two tasks, the first is object detection and the second is image classification, and showed excellent results in detection accuracy and speed, as the optimization method was applied to different architectures of the YOLOv8n model, which led to reducing non-contributing parameters and increasing detection speed. The results of the thesis indicate that the proposed Pruned-YOLOv8n model maintains high accuracy (83%) with a slight decrease of 0.8 compared to the baseline model, while reducing the size of the proposed model by 33% and increasing the detection speed to 196 fps.