The data from COVID-19 has led us to develop a deep learning system for detecting face masks with high efficiency and accuracy. Our model detects whether a person is wearing a mask based on convolutional neural networks (CNNs) and VGG-16 architecture. In order to enhance the model's accuracy and computational efficiency, we used transfer learning with the VGG-16, which took advantage of pre-trained layers to classify images with various characteristics. Various datasets containing images of people wearing masks and those without were used to fine-tune the model, resulting in reliable mask and face detection. The methods were evaluated extensively using benchmark datasets, ensuring reproducibility and comparison with existing ones. Based on the proposed method, accurate results can be achieved in complex, real-world scenarios that may include occlusions and diverse lighting conditions with a short inference time and efficient memory usage.