Advancements and recent trends in Emotion Recognition using facial image analysis and machine learning models

As the demand for systems with human computer interaction grows, automated systems with human gesture and emotion recognition capabilities are the need of the hour. Emotions are understood by textual, vocal, and verbal expression data. Facial imagery also provides a constructive option to interpret and analyse human emotional issues. This paper describes the recent advancements in methods and techniques used to gauge the five primary emotions or moods frequently captured on images containing the human face: surprise, happiness, disgust, normality, drowsiness, through automated machinery. Looking at the recent developments in facial expression recognition techniques, the focus is on artificial neural networks and Support Vector Machine (SVM) in emotion classification. The technique first analyses the information conveyed by the facial regions of the eye and mouth into a merged new image and using it as an input to a feed forward neural network trained by back propagation. The second method showcases the use of Oriented Fast and Rotated (ORB) on a single frame of imagery to extract texture information, and the classification is completed using SVM. The special case of drowsiness detection systems using facial imagery by pattern classification, as automated drowsiness detection promises to play a revolutionary role in preventing road fatalities due to lethargic symptoms in drivers is also discussed.