A master's thesis at the College of Engineering at the University of Basrah discussed redefining the person using neural networks.
The thesis presented by student Sondos Odeh Abdel Hussein discussed methods of re-identifying people through images, video and real-time using graph convolutional neural network (GCNN) and YOLOv5. GCNN has been adopted in a paradigm called Omni-Scale Network (OSNET) to extract network features based on several well-known datasets for human re-identification such as Market1501, GRID, PRID, Dukemtmcvidreid, MARS as well as two custom datasets called SAli and SAli2 which have made sufficient progress in Re -Id compared to other Re-Id datasets. This study also adopted a new method based on YOLOv5 for the first time ever As a real-time detection and identification tool for personal photos and videos.

 The thesis concluded that the adopted method was superior in revealing the identity of individuals compared to previous literature. In addition, the designed system is validated in real time using single and multiple cameras based on YOLOv5 which achieved good performance in terms of accuracy and inference time of the target person.