FACE VERIFICATION WITH OCCLUSION USING CONVOLUTIONAL NEURAL NETWORK (CNN)
Daftar Isi:
- The face is used as an object for identification and verification of a person's identity. Occlusion in the form of glasses has the potential to influence the verification process carried out by the software. Therefore, the software will be built to do the face verification process with occlusion using the Convolutional Neural Network (CNN) method. The CNN architecture used in this study is VGG16 consisting of 13 convolutional layers, 5 pooling layers, and 2 fully connected layers. The dataset used for the training and testing process is divided into 2 types namely the AT&T face database and the primary dataset. The accuracy generated from face verification software with occlusion using primary dataset is 99.13% while the accuracy generated from face verification software with occlusion uses secondary dataset of 95% and the response time generated by 0.2 seconds is approaching realtime.