Jurnal Teknologi Terpadu
Vol. 9 No. 1 (2023): Juli, 2023

Pengembangan Aplikasi Emoticon Recognition dan Facial Recognition menggunakan Algoritma Local Binary Pattern Histogram (LBPH) dan Convolutional Neural Network (CNN)

Haeruddin Haeruddin (Universitas Internasional Batam)
Herman Herman (Universitas Internasional Batam)
Patrick Pratama Hendri (Universitas Internasional Batam)



Article Info

Publish Date
04 Jul 2023

Abstract

In the current modern era facial recognition technology can be found inside of everyday life, but said technology still has a big problem which is deepfake, where in which a deepfake can bypass security systems created with facial recognition as its base, one facial aspect that a deepfake cannot replicate perfectly is the emotion that can be observed from expression, which is why an emotion can be used as a tool to detect a deepfake, which is why an application that can detect both face and emotion at the same time is needed to add security to facial recognition technology, writer has succeeded in creating an application that can do both emotion recognition and facial recognition at the same time using LBPH (Local Binary Pattern Histogram) algorithm and purposive sampling technique for the facial recognition aspect with 67.5% accuracy and CNN (Convolutional Neural Network) algorithm using FER2013 (Facial emotion Recognition 2013) dataset for the emotion recognition aspect with 58.4% accuracy, with CRISP-DM method that can achieve the average accuracy rate of 63%, because currently not many research combine facial recognition using LBPH (Local Binary Pattern Histogram) algorithm and emotion recognition using CNN (Convolutional Neural Network) algorithm at the same time.

Copyrights © 2023