Simultaneous Prediction of Nationality and Gender from Facial Images Using Deep Learning

In this study, a large dataset comprising ۲۰۰۰ images extracted from Wikipedia was used to classify individuals' gender and nationality. For this purpose, two popular deep learning models, namely VGG۱۶ and a custom Convolutional Neural Network (CNN), were trained using the SGD optimizer with momentum. The results from evaluating the models on the training and validation data indicate that the VGG۱۶ model significantly outperformed the custom CNN model. VGG۱۶ achieved ۹۸% and ۹۹% accuracy in classifying nationality and gender on the training data, and ۹۰% and ۹۱% accuracy on the validation data, respectively. This study demonstrates that deep learning models, particularly VGG۱۶, have a high potential for performing complex image classification tasks, including gender and nationality recognition.

کلیدواژه ها:

Gender classification ، Nationality prediction ، Deep learning ، Convolutional Neural Networks (CNN) ، VGG۱۶