Learners mood detection using Convolutional Neural Network (CNN)

摘要

This research concerns about classroom learners mood detection in learning process which is believed to be an important thing to increase learning process effectiveness. Convolutional Neural Network (CNN), a branch of deep learning architectures and a part of Machine Learning, was used as a method in this research. The experiments were conducted through several stages such as face detection, image improvement and model formation. There are 660 images used as training data and the classification process result showed a good result. The accuracy average result was considered as a good result by using 4 layers of CNN i.e. 2 convolutional layers and 2 subsampling layers. Based on the experiments result, the system needs to be developed further by adding more specific data class and training data.