Facial expression recognition using convolutional neural network based approach for 3D facial images

Abstract

Facial expression is defined as the activity of recognizing the internal newlinefeeling or emotional state. It is one of the powerful tools for effective newlinecommunication between persons. They are the cues that control our newlineinteractions with other persons in our vicinity. Thus, they play an important newlinerole in building successful relationships. Facial Expression Recognition newlinesystem attempts to extract features from facial images, classify the newlineexpressions by analyzing the changes in facial features. This system plays a newlinevital role in improving human computer interaction. Most of the works in facial expression recognition depends heavily on the prominent facial landmarks and a reference facial model. Locating the most prominent facial landmarks become tedious because of the complexity of facial musculature and it also needs manual intervention for better accuracy. Model based approaches need to establish a reference model and newlinecomplex functions for mapping which takes intense computation time. So, newlinethis research work provides a new dimension to deal with the above issues by newlineproposing a Convolutional Neural Network based approach that does not rely newlineon the landmarks or a reference model. newline newline