Strategies to Improve the Interactivity of Online Education by Integrating Artificial Intelligence and Biometric Recognition Technology in Educational Technology

Jun He
Faculty of Education, Shaanxi Normal University, Xi'an, 710000, China.
Lixia Li
School of Educational Technology，Northwest Normal University, Lanzhou,730100, China

Abstract:

In educational technology, with the rapid development of information technology, online education has gradually become an important form of modern education. In online education, traditional teaching strategies mostly adopt a unified interactive method, which is difficult to make real-time personalized adjustments according to students' learning status, and it is difficult to monitor students' concentration, resulting in poor interactivity. This paper combines facial expressions and eye features in biometrics, takes advantage of the spatial and temporal features extracted by the multi-task ResNet50-BiLSTM hybrid model, and uses the dynamic strategy adjustment characteristics of D3QN to improve the interactivity of online education. Based on preprocessing, the study uses the ResNet50 model to extract spatial features of facial expressions and eye features, respectively, and sends them to the BiLSTM model for temporal feature capture. The extracted facial expressions and eye features are then fed into the multimodal transformer and interactive attention for feature fusion to capture the intrinsic correlation between different modalities. Finally, the D3QN model is used to adjust the personalized teaching strategy in real time based on the student's situation. The experiment was based on real-time data collected from a school's online education platform. The results showed that after real-time strategy adjustment of the D3QN model that integrates ResNet50-BiLSTM and facial expressions and eye features, the average number of interactions increased by 2.5 times, and the recognition accuracy of ResNet50-BiLSTM on facial expressions reached 0.95, which was 0.04 higher than that of VGG19-BiLSTM. The results show that combining artificial intelligence and biometric recognition technology can effectively improve the interactivity of online education and optimize the overall teaching strategy in real-time.