IoT-based multimodal learning framework for predicting student engagement in English education

With the rapid development of Internet of Things (IoT) and Artificial Intelligence (AI), multimodal learning has become an important part of intelligent systems. It helps improve decision-making by integrating and analyzing various data sources. These technologies have the potential to revolutionize the field of education, especially in smart cities, by monitoring and optimizing the way students are engaged. This study introduces a multimodal learning framework based on Transformers. The framework integrates IoT devices, artificial intelligence and smart city infrastructure and uses optical sensing technology to predict student engagement in English language learning. The framework uses optical IoT sensors to collect real-time, high-fidelity multimodal data. This data includes voice signals, physiological indicators and facial expressions. To ensure accuracy and reliability in dynamic environments, these data are preprocessed and synchronized. In the converter architecture, the dynamic attention mechanism can be adjusted according to the priority of modal features. This helps to address the issues of heterogeneous data fusion, noise immunity and computational efficiency. Experimental results from real classroom environments in smart cities show that the framework achieves a prediction accuracy of 91.2% with an F1 score of 0.90, which is more accurate than the traditional model based on the LSTM method and CNN. In addition, model compression techniques such as quantization and weight pruning improve the computational performance and real-time processing with an average inference time of 10ms. These findings suggest that the framework can enhance intelligent educational systems through accurate and efficient engagement prediction. The proposed strategies show great potential for improving personalized learning and adaptive teaching strategies. These strategies provide scalable and practical solutions for real-world applications in smart cities.