Comparative Study on the Analysis of the Performance of Transfer Learning and the Customized Convolutional Neural Network to Detect Eyewear

In this paper, we propose a method for eyewear detection using Convolutional Neural Networks (CNNs). The objective of the study is to accurately classify images into two classes: “Glasses” and “No Glasses The proposed approach is evaluated on a dataset comprising a total of 5000 images, with a training set consisting of 1947 images in the “Glasses” class and 1554 images in the “No Glasses” class. The testing set includes 834 images in the “Glasses” class and 665 images in the “No Glasses” class.We evaluate the eyewear detection performance using various evaluation metrics, including accuracy, precision, and recall. The outcomes demonstrate how well the suggested CNN-based technique for eyeglasses detection works. The trained model achieves an overall accuracy of 99.17%, with a precision of 0.76% and a recall of 0.44% on the testing set. The discovery advances the field of computer vision and creates new opportunities for improving eyeglasses detection methods through further additional study.