TY - JOUR AU - Yamin, AU - Khin Khin Oo, AU - Htay, Moe Moe PY - 2020/04/13 Y2 - 2024/03/29 TI - Neural Network-Based Expression Recognition System for Static Facial Images JF - International Journal of Computer (IJC) JA - IJC VL - 38 IS - 1 SE - Articles DO - UR - https://ijcjournal.org/index.php/InternationalJournalOfComputer/article/view/1593 SP - 15-21 AB - <p>Affective Computing is a field of studying the human effect to interpret, recognize, process, and simulate in computer science, psychology, and cognitive science. Humans express their emotions in a variety of ways such as body gesture, word, vocal, and mainly facial expression. Non-verbal behavior is a significant component of communication, and facial expressions of emotions are the most important complex signal. Facial Expression Recognition (FER) is an interesting and challenging task in artificial intelligence. FER system in the study three steps including preprocessing, feature extraction and expression classification. In the paper, comparative analysis of expression recognition is implemented based on Neural Network (NN) with three feature extraction methods of Sobel Edge, Histogram of Oriented Gradient and Local Binary Pattern. NN-based expression recognition system achieves an accuracy of 95.82% and 97.68% for JAFFE and CK+ dataset respectively. The result has shown that the Edge features are the effected features for recognizing human expression using still images.</p> ER -