Face detection is arguably one of the most researched aspects of computer vision today. The rapid developments being made in Artificial Intelligence and deep learning technologies which perfects face detection and analysis are spearheading progress in this area. The availability of simple and complex datasets facilitates such progress since they allow for the training and refinement of a model’s performance under various conditions. In addition, the precision and effectiveness of face detection technologies have made them fundamental in various fields such as security, marketing, and healthcare. This research aims to automate human face recognition systems using deep learning algorithms and pre-trained models such as AlexNet, GoogleNet, ResNet50, and VGG16. This research attempts applying these models to specific datasets, training them to better suit the data used through model adjustment, also referred to as “fine-tuning”. Moreover, this research seeks to evaluate these models on face recognition to benchmark the accuracy and efficiency metrics, ultimately finding the model best suited for operations in exigent conditions demanding rapid and precise recognition. The results showed that the AlexNet model was the most efficient among those tested for human face recognition, outperforming GoogleNet, ResNet50, and VGG16 with the highest performance metrics. With AlexNet, classification accuracy achieved about 94.56%, with total accuracy at 96.62% and overall positive recall at 93.25%. These results demonstrate the AlexNet model’s effectiveness in accurately extracting and processing vital facial features, making it the best model among those used in this study.