Emotion Classification of Facial Images Using Machine Learning Models

Humans find it very easy to detect emotions through facial expressions. It is a herculean task for the machine to replicate the same. But recently Machine Learning (ML) is coming to the aid of computer vision in detecting facial emotions accurately from digital photos. In this work a system that uses machine learning algorithms is designed for identifying the emotions based on captured digital images of faces. A total of eight ML models has been designed to classify the facial images into an emotion. Human emotions mainly include happy, sad, surprise, anger, serious and disappointment. ML algorithms like K-Nearest Neighbour (KNN), Naive Bayes (NB), Decision Tree (DT), Random Forest (RF), Gradient Boosting (GB), AdaBoost Classifier, Histogram Gradient Boosting Classifier (HGB), Linear Support Vector Classification (SVC) were used in designing the models. In this research, the HGB based model attained maximum accuracy of 75.67%. Facial Emotion Classification (FEC) is significant in areas like customer feedback, surveillance, mental disease diagnosis, Human Computer Interaction and social behavioural analysis.