The project aims to detect deepfake images. I used tensorflow and keras to build the model. Inorder to train the model efficiently, I used early stopping and model checkpointing. I also used data augmentation[this was done locally] to increase the size of the dataset.
The dataset and problem statement was taken from the following link: https://iplab.dmi.unict.it/deepfakechallenge/
Only Task-1 of the problem statement is implemented from the link above.