Title Page
Abstract
국문 초록
Contents
Abbreviations 10
Chapter 1. Introduction 13
1.1. Purpose and Motivations 14
1.2. Development of Online Learning systems 15
1.3. The limitations of online education 16
1.4. Problem Statement 16
1.5. Research Objectives 17
1.6. Related works 17
Chapter 2. Literature review 18
2.1. Deep learning for Computer vision 18
2.2. Convolutional Neural Networks (CNN) 19
2.2.1. Feature detection 20
2.2.2. Classification 20
2.3. Overview of Facial expression 21
Chapter 3. Methods and Datasets 23
3.1. Customized Algorithm 24
3.1.1. Preprocessing 24
3.1.2. Proposed CNN Architecture 26
3.1.3. Classification of facial expressions 27
3.2. Datasets 28
3.2.1. FER2013 28
Chapter 4. Results 30
4.1. Test the model 30
4.1.1. Data preparation 30
4.1.2. Training Parameters 31
4.2. Discussion and comparison 31
4.3. Application 32
4.3.1. APP idea and documentation 33
4.3.2. Result of APP 33
Chapter 5. Conclusion and Future works 35
5.1. Conclusion 35
5.2. Future works 36
References 37
Table 1. An overview of the facial expression datasets. 28
Table 2. Experimental environment. 30
Figure 1. Diagram of AI including DL and ML 18
Figure 2. Architecture of CNN 19
Figure 3. Plutchik's wheel of expressions 22
Figure 4. Workflow visualization 23
Figure 5. Customized Algorithm 24
Figure 6. Data preprocessing 25
Figure 7. CNN layers 26
Figure 8. CNN architecture 27
Figure 9. Samples of FER2013 dataset 29
Figure 10. Loss/ Accuracy plot for the model with a learning rate of 0.001 for 60 epochs 31
Figure 11. Loss/ Accuracy plot for the model with a learning rate of 0.0001 for 60 epochs 32
Figure 12. Different emotion states 34
Figure 13. Graphical view of emotion and time. 34