<![CDATA[Understanding students’ emotional involvement is essential for optimizing teaching strategies and improving learning outcomes in the classroom. This study employed a Research and Development (R&D) method aimed at developing a system to identify students’ learning interest based on real-time facial expression detection. The system consists of three main components: input in the form of facial images captured by a webcam, expression classification processed by a laptop, and output in the form of information indicating the level of student interest. The devices used in testing included a Logitech C310 webcam connected to an MSI Cyborg 15 A12VE-074ID laptop, equipped with an Nvidia RTX 4050 GPU. The algorithm used for the recognition and classification process was a Convolutional Neural Network (CNN) implemented through the YOLO version 11 framework. The dataset consisted of 1,128 student facial expression images classified into two main categories: (1) interested expressions, including happy and surprised, and (2) uninterested expressions, including angry, fearful, sad, and disgusted. These expressions represent basic human emotions. The test results showed that the system achieved an accuracy rate of 86.04% and a precision rate of 89.57%, with an average detection time ranging from 31 to 50 milliseconds. Testing conducted over five days of classroom sessions demonstrated that the system was able to operate stably and accurately using a webcam, even under varying lighting conditions, angles, and facial movements, with an average frame rate of 30–32 FPS.]]>
