Facial Expression Classification Using Vanilla Convolution Neural Network

Abstract

Automatic Facial Expression Recognition (FER) is an active research area in artificial intelligence and image processing. One of the main challenges in FER is the automatic feature extraction. In this paper, the proposed 10 -layer Convolutional Neural Network (CNN) architecture can automatically detect important features without human supervision. This research uses Extended Cohn-Kanade (CK+) dataset and Japanese Female Facial Expression (JAFFE) dataset for facial expression recognition. The proposed architecture achieved an accuracy of 99.3% and 78.1% in classifying facial expressions in CK+ and JAFFE datasets.