Face emotion recognition (FER) plays a crucial role in the field of human-computer interactions, as it holds a substantial significance in psychology analysis and effective communication. This research area has been attracting many interests to improve the accuracy of emotion classification and to resolve the challenges of the task. This study investigates how deep learning is capable of recognizing emotion status of students’ face images. The study shows the pipeline from the data preprocessing to deep neural network learning models, namely, the learning models are convolutional neural network (CNN) and deep Belief network (DBN) for recognizing emotion expression of students’ videos/images. This study shows the performance and comparison between CNN, DBN and their hybrid. The result demonstrates that CNN and DBN can address well the emotion recognition problem.