Klasifikasi Gender Melalui Citra Mata Menggunakan Metode Convolutional Neural Network (CNN) dengan Model Arsitektur VGG-19

Gender is not just a biological difference between men and women, but also reflects roles, characteristics and identities that are influenced by social and cultural factors. One way to identify gender is through eye image analysis, as eye patterns and structures can have distinctive characteristics between men and women. In this study, gender classification based on eye image is performed by utilising Convolutional Neural Network (CNN), which is one of the deep learning methods designed to recognise patterns in visual data. The model used in this study is VGG19, a CNN architecture that has more layers than previous models such as VGG16. The dataset used consists of 11,525 eye images, which are divided into 6,323 male eye images and 5,202 female eye images. The experimental results show that the VGG19 model is able to achieve an accuracy of 91.33%, which is higher than previous studies using other CNN architectures. This finding shows that VGG19 has a good ability to recognise unique patterns in eye images, thus improving the accuracy in gender classification. This success opens up opportunities for further research in various fields, such as biometric systems, security, and more accurate and efficient image-based pattern recognition technology.