Image Processing: Human Facial Expression Identification using Convolutional Neural Networks

Abstract

Facial expressions identification is the growing area of interest as it attracts the advancement in technology by focusing on human computer interaction. Many researchers has got their hands on various approaches for automatically generating the expressions of human faces. Emotions and expressions are inter relatable where expressions are facial movement for expressing the emotions of human being. Image processing is the technology which helps in identifying expressions by including factors such as the face detection, the feature extraction and the expressions classification. Two datasets FER-2013 and CK+48 are used in the process of identifying the expressions like sad, fear, happy, angry, surprise, neutral, disgust, contempt.  HAAR features based adaboost cascades are used to identify the features of a face which helps in detecting facial points and makes it easy for further process of expression detection. The deep learning technique, convolutional neural network (CNN) is implemented for classification of expressions for prediction