%0 Thesis
%9 S2
%A Putra, Werda Buana
%A Arifin, Fatchul
%B Pendidikan Teknik Elektronika dan Informatika
%D 2019
%F UNY:66399
%I Program Pascasarjana
%K Depression, convolutional neural network, Euclidian distance, FURIA fuzzy
%T A Facial Expression Recognition System to Monitor Student’s Mood In A Classroom.
%U http://eprints.uny.ac.id/66399/
%X Depression is a commonly unattended health problem, unbounded by age border, and greatly affecting student’s performance in their study. To prevent it, the writer built a real-time facial emotion recognition system, so the teacher can monitor students’ mood through class activity. The system should be reliable enough when running on mid-end computer specification.  The writer use transfer learning for the dataset, pre-trained by utilizing convolutional neural network theory. The system use Euclidian Distance as the basis to do the facial landmark, and applying FURIA fuzzy rules to calculate and get the desired facial emotion result. JAFFE image set will be used to test the system’s accuracy by comparing the result shown by the system and the already expert-arranged image set. The student will be given questionnaire to measure their stress. The questionnaire result will be used to analyze whether the use of the system able to reduce student’s stress or not.  The system built is able to properly classifying 7 human facial emotion captured by webcam. By comparing the result of the emotion recognition processing and the expert- arranged emotion on JAFFE image set, it is concluded that the system’s accuracy reached 90%. The result from the questionnaire shows that the use of the system able to detect student’s mood early so the teacher may minimize student’s stress.