Step 1. Install libraries
conda env create -f environment.yml
Step 2. Download processed fer2013 CSV files from this link
https://1drv.ms/u/s!AmeTT2EpSz40hRbRieQ2kiQGLQ20?e=I27fR1
The original dataset can be found from this link
https://www.kaggle.com/competitions/challenges-in-representation-learning-facial-expression-recognition-challenge/data
Step 3. Put the processed fer2013 CSV files in the following folder
data
Step 4. Download pre-trained models from this link
https://1drv.ms/u/s!AmeTT2EpSz40hFYvpEYq3gvIUTE5?e=XgTXwb
Step 5. Put the pre-trained models in the following folder
checkpoint
To train the model from scratch, run the following
python main_fer2013.py
After training the new model will be saved at the checkpoint folder
For evaluation, edit the cm_resmasking file to use the model you want and run the following
python cm_resmasking.py
After running the above file,the Confusion matrice will be saved at the cm folder