As advanced technologies involving multimodal AI sensors have been increasingly adopted to improve teaching quality, student engagement, and learning outcomes, it is crucial to evaluate their effectiveness in education. In this study, we surveyed 103 participants to evaluate the technologies in terms of effectiveness and user perception by analyzing the obtained statistics. The perception and use of multimodal AI sensors were significantly correlated with student engagement while they were less correlated with learning outcomes, as various factors might influence academic achievement represented by learning outcomes. The results of the survey also showed that educational institutions need to prioritize ethical considerations regarding privacy protection and data usage, especially data containing personal information, and persuade stakeholders in education to use innovative devices such as multimodal AI in the classroom. Then, various AI technologies can be more effectively and efficiently used in education.