Comparative Study between Machine Learning and Deep Learning in Facial Emotion Recognition Application Design Designing Facial Analyzing and Emotion Recognition System Using Convolutional Neural Networks (CNN)

  • 07 Dec 2020
  • Ongoing Research - Informatics & Communication

Dr. Tarek Barhoum; Khloud Al Jallad; Rahma AbuZafra; Lana Abdullah; Rasha Albizreh and Rouaa Alaraj

Researchers

Post Graduate Studies, Research & IR Council Meeting No. 1, 22/11/2020

Date of Acceptance


Phase one: make a comparative study between machine learning algorithms and deep learning algorithms in facial emotion recognition field by designing interactive android software that classifies five fundamental emotions: happiness, nature, sadness, surprised, and anger from an image that sent from person to another person. In addition to that many experiments will be done with goal to get an accurate comparative study. Also many types of image datasets for facial expression recognition will also be used and tested. Also research aims to build an enhanced database in order to reach the best performance in the algorithms used in both types of learning systems.

Phase two: build interactive system that classifies five fundamental emotions: happiness, fear, sadness, disgust and anger from an image. Many experiments will be done using deep learning approaches by building new image dataset for facial analyzing and emotion recognition. The system will work on face detection with deep learning model to extract facial's features then obtain correct classification.

Abstract