- python >= v3.8
- tensorflow >= 2.4
- opencv = v4.5.3
- PIL: pillow = v7.2.0
- Cuda : for the use of GPU for faster training time
- pip install tensorflow
- pip install opencv-python OR pip install opencv-contrib-python [for additional opencv community files]
- pip install Pillow
original dataset from https://www.kaggle.com/msambare/fer2013
- Folders: emojis, models, test, train are in root directory
- cnn_train and gui are python codes that also need to be in the root directory
- emojis: images for each class of emotion that a model has to predict
- models: trained neural network model
- test: test dataset
- train: training dataset