Early Fusion of CNN + BoVW Features for Facial Emotion Recognition on FER2013 without additional training data

Simple item page
dc.contributor.advisorDr. Shazzad Hosain
dc.contributor.authorMinhazul Abedin Toshin
dc.contributor.authorNashrah Haque
dc.contributor.authorMd. Rifat Bin Yusuf
dc.contributor.authorLabib Rahman
dc.contributor.id1931672042
dc.contributor.id1931857042
dc.contributor.id1912217042
dc.contributor.id1931740042
dc.coverage.departmentElectrical and Computer Engineering
dc.date.accessioned2025
dc.date.accessioned2025-07-21T10:00:18Z
dc.date.available2025-07-21T10:00:18Z
dc.date.issued2023
dc.description.abstractFacial emotion recognition is an important research area with various applications such as human-computer interaction, psychology and social robotics. Consequently, there has been active research in the field using the approach of Convolutional Neural Networks (CNNs), for feature extraction and inference. In our paper, we explored the use of Bag of Visual Words(BoVW) and CNNs on the FER2013 dataset without any additional training data. To combine the features from both methods, we concatenated the feature vectors obtained from BoVW and CNN. Subsequently, we employed a Support Vector Machine (SVM) classifier to train and classify the concatenated feature vectors. The evaluation of our approach on the FER2013 dataset yielded an accuracy of 62%. Although this accuracy level indicates room for improvement, it demonstrates the potential of utilizing both BoVW and CNN in facial emotion recognition tasks. Overall, this study showcases the effectiveness of combining the BoVW approach with CNN for facial emotion recognition. The results obtained serve as a foundation for further investigations and advancements in this field.
dc.description.degreeUndergraduate
dc.identifier.cd600000199
dc.identifier.print-thesisTo be assigned
dc.identifier.urihttps://repository.northsouth.edu/handle/123456789/1300
dc.language.isoen
dc.publisherNorth South University
dc.rights©NSU Library
dc.titleEarly Fusion of CNN + BoVW Features for Facial Emotion Recognition on FER2013 without additional training data
oaire.citation.endPage61
oaire.citation.startPage1
Files
Original bundle
Now showing 1 - 2 of 2
Thumbnail Image
Name:
600000199.Abstract.pdf
Size:
221.43 KB
Format:
Adobe Portable Document Format
Description:
Download
Thumbnail Image
Name:
600000199.pdf
Size:
474.98 KB
Format:
Adobe Portable Document Format
Description:
Download
License bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
license.txt
Size:
1.93 KB
Format:
Item-specific license agreed to upon submission
Description:
Download
Collections
Theses - Undergraduate