Identification of Fatigue from Facial Expressions Using Transfer Learning

Initially, teaching and learning activities were carried out face-to-face in the provided room, but now they have switched to online. Online learning has an impact on student learning disengagement, which is known through indicators of aspects of emotional exhaustion, physical fatigue, cognitive fatigue, and loss of motivation. Besides, the teacher must provide the material that has been provided. The teacher must also pay attention to all students who are participating in the online learning. This can be overcome by a system that can detect student disengagement using a camera device. The system works by scanning the direction of students' faces and views using OpenCV technology and Transfer Learning methods. Using context, facial expressions, and heart rate can be used to recognize student disengagement. However, with the widespread availability of cameras, it is easier to identify disengagement using facial expressions. The facial expression recognition system in this study will use the FER2013 dataset and Transfer Learning method. Facial expression recognition using the FER-2013 dataset and Transfer Learning method has a reading accuracy rate of 68% in 25 epochs. Then, after being implemented as an impression parameter in the disengagement identification system using 7 scenarios, the accuracy rate is 83.33%, precision is 100%, recall is 75%, and the f1-score is 85.71%.