Online education has seen rapid development, especially due to its unparalleled convenience. Recent circumstances, such as the COVID-19 pandemic, have forced many schools worldwide to delay opening and adopt online education as a primary teaching method. However, the efficiency of online classes has been a subject of skepticism. Compared to traditional face-to-face classes, online classes often lack direct, timely, and effective communication and feedback between teachers and students. To address this challenge, we present a model that can help improve the online learning experience.
Our approach utilizes cameras in devices to collect students' facial images during online classes. These facial images are then analyzed using a compact deep learning model based on the architecture of Convolutional Neural Networks (CNN). The model classifies the facial expressions into seven categories of emotions: Angry, Disgust, Fear, Happy, Neutral, Sad, and Surprise.
By storing and analyzing this emotion data in our database, we aim to offer valuable insights to online school platforms and educators. The collected emotion data will be represented in the form of histograms, providing an overview of students' emotional states during the online learning process. With this information, teachers can adjust their teaching strategies accordingly, tailoring their approach to improve the overall efficiency of online teaching.
Enhanced Understanding: By understanding the emotional states of students during online classes, educators can gauge their engagement and identify potential areas of improvement.
Personalized Approach: The emotion analysis results will allow teachers to personalize their interactions and teaching methods, catering to individual emotional needs.
Real-time Feedback: Although online classes lack immediate face-to-face communication, emotion analysis can offer a form of real-time feedback to bridge this gap.
Optimized Learning Environment: With insights gained from the emotion analysis, the online school platforms can further optimize the learning environment for students' benefit.