Deep Convolution Neural Network for Facial Expression Recognition

Authors

  • Soleen Basim Mohammed , Adnan Mohsin Abdulazeez

Abstract

The Facial Expression Recognition (FER) is still considered as an open research problem and proposing new techniques for more accurate recognition is challenging task. Thus, the main aim of this paper is a proposed algorithm for facial expression recognition FER. The proposed method is based on deep neural network, namely, Convolutional Neural Network (CNN) technique. The main objective of the proposed study is classifying emotions of facial expression into different types, natural, smile, surprise, disgust, squint and scream emotions are used in this study. The efficiency of the selected deep CNN for feature extraction and classification has been proven comparing other techniques. For feature extraction Principle Component Analysis (PCA) method. Experiments are carried out on the Extended Cohn-Kanada (CK+) and the Japanese Female Facial Expression (JAFFE) datasets to show the effectiveness of the proposed method. The evaluation results obtained height recognition rate 98.5%.

 

Downloads

Download data is not yet available.

Downloads

Published

2021-02-05

How to Cite

Soleen Basim Mohammed , Adnan Mohsin Abdulazeez. (2021). Deep Convolution Neural Network for Facial Expression Recognition. PalArch’s Journal of Archaeology of Egypt / Egyptology, 18(4), 3578-3586. Retrieved from https://archives.palarch.nl/index.php/jae/article/view/6874