Real Time Facial Expression Recognition Based On Deep Neural Network

Authors

  • T. Ambikadevi Amma Professor & Principal, Department of Computer Science and Engineering, Nehru College of Engineering and Research Centre, Pampady, India
  • M. R. Sruthy Assistant Professor, Department of Computer Science and Engineering, Nehru College of Engineering and Research Centre, Pampady, India
  • S. Divya Assistant Professor, Department of Computer Science and Engineering, Nehru College of Engineering and Research Centre, Pampady, India
  • P. Renuka PG Scholar, Department of Computer Science and Engineering, Nehru College of Engineering and Research Centre, Pampady, India

Keywords:

Artificial Intelligence (AI), Convolutional Neural Network (CNN), Emotion recognition, Facial expression recognition (FER), YOLOv2

Abstract

Now-a-days with the continued development of artificial intelligence facial emotion recognition has become more popular. The emotion recognition plays a major role in interaction technology. In interaction technology the verbal components only play a one third of communication and the non-verbal components plays a two third of communication. A facial emotion recognition (FER) method is used for detecting facial expressions. Facial expression plays a major role in expressing what a person feels and it expresses inner feeling and his or her mental situation or human perspective. This paper aims to identify basic human emotions with the combination of gender classification and age estimation. The facial emotions such as happy, sad, angry, fear, surprised,
neutral emotions are considered as basic emotions. Here proposes a real time facial emotion recognition system based on You Look Only Once (YOLO) version 2 architecture and a squeezenet architecture. The yolo architecture is a real time object detection system. Here it used for identify and detect faces in real time. These images are captured by using anchor boxes for accuracy. The second architecture is squeezenet and is used for gender classification and age estimation. It provides significant, accurate object detection and extracts high-level features that help to achieve tremendous performance to classify the image and detecting objects. Both the architectures provide accurate result than other methods with the large no of hidden layers and cross validation in the neural network.

Downloads

Download data is not yet available.

Downloads

Published

14-07-2020

How to Cite

[1]
T. A. . Amma, M. R. . Sruthy, S. . Divya, and P. . Renuka, “Real Time Facial Expression Recognition Based On Deep Neural Network”, IJRESM, vol. 3, no. 7, pp. 59–63, Jul. 2020.

Issue

Section

Articles