Skip to content

Files

Latest commit

 

History

History
18 lines (9 loc) · 756 Bytes

README.md

File metadata and controls

18 lines (9 loc) · 756 Bytes

fer-exp

The FER2013 and AffectNet datasets must be downloaded from:

https://www.kaggle.com/datasets/msambare/fer2013

http://mohammadmahoor.com/affectnet/

and must be pasted in fer2013 and affectNet folders respectively.

You can download the respective models from:

https://drive.google.com/drive/folders/1jmnVLTgpXZOYN-tWYOxkrI-VqghfNAX9?usp=sharing

and paste them into the checkpoint folder

In this project, we compare the performance of four deep learning models - our own CNN model, VGG-16, ResNet-50, and DenseNet-121 - on the FER-2013 dataset for facial emotion recognition. This research provides insights for future work in the area of facial emotion recognition and can be useful in developing more accurate and efficient models.