Facial Expression Detection and Recognition using Geometry Maps

Abstract

Processing in faces is done differently compared to other stimuli, because of challenges in differentiating among highly familiar stimuli to recognize the individual since there is a necessity of the society to do it successfully. Since there can be vast growth in the field of machine learning, humans from past few years trying to make computer interaction with the humans. To do so it took various steps whether it is about the facial expression recognition or about other fields of machine learning. We proposed a novel way to deal with distinguishing the feeling in view of the lips structure over the time frame. This paper presents the work done in the field of facial expression recognition with the help of geometric maps representation of the lips portion including the Support Vector Machine (SVM) classifier to get more accuracy in the final result and then plot a graph which gives us an approximate estimation of the expression of the humans. The experiment is done on Cohn-Kanade database and obtained 94% accuracy.