Real-time Facial Emotion Recognition with Deep Neural Networks

Main Article Content

Samad Azimi Abriz , Majid Meghdadi


Face emotions are an important part of human communications that helps to perceive others’ intentions and behaviours. This crucial element creates a connection and interpersonal communications can be apprehended using this factor. So, there is no wonder that numerous researches have been dedicated to this matter during recent decades and facial emotion recognition is a critical matter in computer vision and artificial intelligence. One of the challenges of AI is to achieve high accuracy and great performance in the model. Fortunately, this has been dealt with by emerging deep learning networks and setting aside traditional machine learning methods. In this study, we established a model called ResEZAP (Residual Extended Zero Average Pooling) of deep learning networks for real-time facial emotion recognition and achieving an acceptable accuracy by reducing computational complexity. In this paper, the FER2013 dataset is used for training and the model accuracy with the test dataset is 69.74.


Download data is not yet available.


Metrics Loading ...

Article Details

How to Cite
Samad Azimi Abriz , Majid Meghdadi. (2021). Real-time Facial Emotion Recognition with Deep Neural Networks. Turkish Journal of Computer and Mathematics Education (TURCOMAT), 12(13), 7546–7557.