CSCE633 class project -- Lighten the Face Forgery Detection Network by using MobileNet
This is our implementation of the CSCE633 course project. After downloading all files, you should first unzip the "original_database" and "extended_database" in the "dataset" file and move these two datasets to the main directory. After that, you can simply use train.py to train a model based on a chosen dataset and use test.py to get the prediction accuracy of a trained model. You can also use roc_curve.py to draw the ROC curve based on a trained model.
The report is also attached.