<![CDATA[This study develops a facial emotion detection model using the VGG-19 architecture on the Convolutional Neural Network (CNN) method. The study aims to improve the accuracy of facial expression recognition in a variety of applications, including the fields of education and marketing. The dataset used consisted of 7 categories of emotions, with 80% of the training data and 20% of the testing data. The research process includes dataset collection, data pre-processing, CNN model design, and model training with epoch variations. The results showed an increase in accuracy as the number of epochs increased, with the highest accuracy reaching 69.81% in training data and 63.30% in validation data after 100 epochs. The model showed good performance in classifying "happy" emotions, but had trouble distinguishing between sad, neutral, scared, and angry emotions. The study proves CNN's effectiveness in classifying facial emotions, although there is still room for improvement, especially in distinguishing similar emotions.]]>
Copyrights © 2024
