ABSTRACT
During this pandemic scenario, most people are getting vaccinated by violating the age rules defined by the. We found a few scenarios where persons aged 30–40 have claimed their age as 45 and got vaccinated during the early days of vaccination. In some scenarios, students aged 18–20 also got vaccinated before the government declared vaccination for those above 18 years. By considering all these scenarios, the proposed system has implemented a model that can recognize the person’s gender and age for verification before allowing them to register for vaccination. Previously, different models could recognize the face along with their age and gender. However, this system proved its efficiency by predicting accurate results with and without masks. This system has achieved an accuracy of “98.51%,” which is far better than the previous models stated in the literature survey. The 130proposed model has utilized the pretrained model best for age and gender are known as “Tal Hassner and Gil Levi.” The existing THGL CNN model consists of five-layered architecture in which three layers are CNN layers, and two layers are dense. In the existing model, first two layers take care of the normalization process with the help of local search algorithms, and the third layer implements 384 filters to extract the necessary features. The first dense layer uses these extracted features and uses 512 neurons to predict the gender of the person using ReLu activation. The last layer uses the soft-max layer because age is a multi-classification process. This model tested the application using both cropped images and over-sampled images, and the average prediction rate is taken into account.
