Implementasi Algoritma Convolutional Neural Network (CNN) Dengan Optimizer Adam Dalam Deteksi Emosionalpada Wajah Manusia

Poppy Amalia (STMIK TIME Medan)
Robby Wijaya (STMIK TIME Medan)
Chandra Chandra (STMIK TIME Medan)
Pieter Octaviandy (STMIK TIME Medan)
Wilson Wilson (STMIK TIME Medan)
David David (STMIK TIME Medan)
Andy Andy (STMIK TIME Medan)
Herman Herman (STMIK TIME Medan)
Edi Edi (STMIK TIME Medan)
Didik Aryanto (STMIK TIME Medan)
Joni Joni (STMIK TIME Medan)
Johanes Terang kita Perangin Angin (STMIK TIME Medan)

JURNAL VOKASI TEKNIK (JUVOTEK)
Vol 3 No 1 (2025): April Issue 2025 Issue | JUVOTEK JURNAL

Emotional detection in human faces is an important area in image processing and emotion understanding. This research develops an emotion detection application using the Haar Cascade Classifier algorithm to detect faces and Convolutional Neural Network (CNN) with the Adam optimizer to analyze emotions. The application developed successfully detected seven types of basic emotions (anger, disgust, fear, happy, neutral, sad and surprised) in real-time with an overall accuracy of 90%. The combination of CNN and Adam optimizer shows good performance with increasing accuracy and consistent decreasing loss as the epoch increases, although there are indications of overfitting. The research results show that this system can be relied on to detect various facial expressions and provides an effective and accurate solution for emotional detection.