“…Before loading input images, the working environment need some preparation such as installing the required libraries, then start to load the dataset. A preprocessing data, is required at this stage including cleansing the dataset from low quality images which will confuse the training model, A relabeling age group images to five categories {(00-10), (11)(12)(13)(14)(15)(16)(17)(18)(19)(20), (21)(22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35), (36-54), (55-90)}, splitting dataset to 80% train, 10% test and 10% validation and finally, set the batch size. The age estimation model is trained with three pre-trained-weight, and two deep learning algorithms (VGG-Face and ResNet50) as it illustrated in the experimental work.…”