• Aanchal Singh
• Jyoti Kansari
• Vivek Kumar Sinha

In computer-aided detection (CAD) Face Recognition is a vital issue, this paper presents a novel and efficient face recognition technique using VGG16 working on the principle of Transfer Learning. JPEG, GIF, and PNG image forms are used to investigate these issues, and they require skill to discern facial activity. This paper essentially centers on designing a CAD model which performs face recognition of 15 subjects, each subject with 11 images per subject, the facial expression include images with glasses, without glasses, happy, sad, sleepy, surprised or wink. The classification problem, used to differentiate among all these different set of categories. The suggested Classification model is motivated by the concept of Transfer Learning, and it employs a deep pre-trained VGG-16 architecture to extract features from GIF images, as well as a deep learning method with a Softmax layer acting as a classifier. The experiment is carried out by comparing two configurations one with normal VGG architecture and other with Customized Fully Connected Layer in VGG architecture. The methodology recorded test accuracy of 83.11% using first customized configuration and 92.80% accuracy with second customized configuration.

Transfer Learning, Face Recognition, FRS, VGG16, GIF, Computer-Aided Detection, Classification Model, Yale dataset.

"Face Recognition Using Transfer Learning by deep VGG16 model", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.9, Issue 4, page no.b121-b127, April-2022, Available :http://www.jetir.org/papers/JETIR2204115.pdf

"Face Recognition Using Transfer Learning by deep VGG16 model", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.9, Issue 4, page no. ppb121-b127, April-2022, Available at : http://www.jetir.org/papers/JETIR2204115.pdf