Abstract:
Biometric identification has become increasingly popular as technology progresses, especially
with regard to facial and ocular recognition. The study investigates the use of deep learning
methods more especially, the VGG-19 model in the analysis of eye pictures with the goal of
identifying nationality. Five different classes representing the perspectives of people from
Bangladesh, Vietnam, South Korea, China, and South Korea are the subject of the study. The
dataset is made up of a small range of carefully selected high resolution eye pictures that
accurately depict each nationality. A deep learning model is trained to identify minute patterns
and features in the eye pictures that differentiate people from the aforementioned ethnicities
using the VGG-19 architecture. By means of extensive testing and optimization, our model
attains a remarkable level of precision.
