NORMA eResearch @NCI Library

Facial Emotion Recognition using Deep Convolutional Neural Network

Langute, Sachin Pralhad (2022) Facial Emotion Recognition using Deep Convolutional Neural Network. Masters thesis, Dublin, National College of Ireland.

[thumbnail of Master of Science]
PDF (Master of Science)
Download (796kB) | Preview
[thumbnail of Configuration manual]
PDF (Configuration manual)
Download (1MB) | Preview


The recognition of facial expressions (FER) is critical for social communication. However, existing research has limitations when it comes to addressing facial expression differences related to variations in the demographics such as age, gender, etc. In face-to face encounters, facial expressions communicate nonverbal information. Since the last three decades, researchers have become increasingly interested in detecting facial expressions automatically, which is very important for Human-Computer interaction. The average person exhibits seven various emotions depending on the scenario, which include anger, sorrow, happiness, surprise, disgust, neutral, and afraid. Every person has their style to show emotions, which cannot be related culturally. Traditional machine learning algorithms can need a sophisticated feature extraction procedure and yield poor results. Artificial Neural Networks (ANN) have been developed to address some of these constraints. The latter produce good results but do not address all the issues such as camera angle, head position, occlusions, and so on. In this research, the author investigates neural network models that are employed in the field of face emotion identification. The author also offers a bilinear pooling-based architecture to build on earlier work's achievements and to give solutions to these reoccurring restrictions. This method vastly increases the performance of designs based on traditional CNNs. This study investigates deep learning strategies for face emotion identification based on Convolutional Neural Networks as well as the VGG16 model. Furthermore, input data is extended by rotation, cropping, and flipping.

Item Type: Thesis (Masters)
Uncontrolled Keywords: Facial emotion recognition; Convolutional Neural Network (CNN); VGG16; Deep Learning; Classification; Machine Learning
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
B Philosophy. Psychology. Religion > Psychology > Emotions
Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions: School of Computing > Master of Science in Data Analytics
Depositing User: Tamara Malone
Date Deposited: 22 Feb 2023 15:25
Last Modified: 02 Mar 2023 09:35

Actions (login required)

View Item View Item