Feature Extraction Face-Off: A Comparative Analysis of InceptionV3 and VGG19 for Face Recognition
Shem L. Gonzales · International Journal of Advanced Research in Science Communication and Technology · 2023
Face Recognition is one of the popular interest of researchers due to its demand in security and authentication. Although traditional approaches such as LBPH and Eigenfaceshave still proven their performance, newly state-of-the-art algorithms are available for application. This study compared the performance of two CNN pre-trained architectures namely: InceptionV3 and VGG19 with an SVM classifier. Besides VGG19’s respectable results, it lags behind InceptionV3 by 3.86% in precision, 3.23% in recall, 3.54% in f1-score, and 1.85% in roc-auc score.