This study explores how mind wandering—a common mental drift away from tasks—can be detected using both brainwave activity (EEG) and facial expressions captured by a regular webcam. The researchers worked with 26 college students who watched educational videos while their EEG signals and facial movements were recorded. Using machine learning, specifically a random forest model, the study tested whether combining EEG data with video features improved the detection of when participants’ attention shifted away from the task. The results showed that combining both types of data led to the best accuracy, with an average AUC of 0.68 for individual-based tests and 0.56 for cross-participant predictions. The study also examined personal differences that influenced detection. Participants with more stable mental states or greater confidence in their self-reports of focus had higher detection accuracy. In other words, people who were more aware of their attention patterns or whose attention fluctuated less were easier to classify correctly. Furthermore, models trained on a larger and more diverse set of participants performed better overall, especially when combining EEG and video data. In short, this work demonstrates that a multimodal approach—merging brain and facial signals—can enhance the accuracy of detecting mind wandering. The findings show promise for real-world applications, especially in online learning environments where maintaining attention is a challenge. By identifying when someone’s focus drifts, such systems could potentially help boost engagement and learning outcomes