DEVELOPMENT OF FACE AND EXPRESSION RECOGNITION SUBSYSTEM IN EXPRESSION IDENTIFICATION SYSTEM IN INTEGRATED SYSTEM SMART CLASSROOM

• Main Author: PRAYOGA NUGRAHA (NIM : 13213034), PRADIPTA
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/29954
• Institution: Institut Teknologi Bandung

<p align="justify"> Nowadays, Indonesia faces difficult problems, especially in education sector. Education in Indonesia ranked 5th in ASEAN and ranked 108th in the world. All this time, Indonesian government focused only on number of graduates. Quality of education has yet to become government’s priority, whereas number of graduates without good quality of education will not affect much on each individual’s quality. To raise our education’s quality, teaching method must be designed so that it can make the students motivated. This is because teaching method mattered 72.6% on student’s motivation. To know whether the method used by teacher is effective or not, there is a need to have a system that can recognize faces and their expressions. In this final project, there will be explanations about design and implementation about development of face and expression recognition subsystem in expression identification system in integrated system smart classroom. This subsystem will be able to recognize faces and the expressions of the students so that teacher may be able to evaluate teaching method based on student’s reactions. This system is implemented using 2 method, which is using viola-jones method continued by CNN, and by using YOLO library. System then tested against distance, angle, and lighting condition. The result is, system using YOLO, along with right dataset, may be able to withstand against various face angles and lighting condition with 100% accuracy. System using YOLO may recognize object up to 5m, but the resolution must be adjusted to minimum of 1250 pixels. <p align="justify">