Ecomended Development Environment

• GPU: RTX 4090(24GB)

• CUDA: 11.3

• PyTorch 1.11.0

• Python 3.8(ubuntu20.04)

1. Dataset

Download datasets and place them in ./data/ folder.
For example, download Celeb-DF-v2 and place it:

.
└── data
    └── Celeb-DF-v2
        ├── Celeb-real
        │       └── *.mp4
        ├── Celeb-synthesis
        │       └── *.mp4
        ├── Youtube-real
        │       └── *.mp4
        └── List_of_testing_videos.txt

For other datasets, please refer to ./data/datasets.md .

2. Pretrained model

We provide weights of EfficientNet-B4 trained on GFADE from FF-raw.
Download and place it in ./weights/ folder.

Inference

For example, run the inference on Celeb-DF-v2:

CUDA_VISIBLE_DEVICES=* python3 src/inference/inference_dataset.py \
-w weights/98_0.9997_val.tar \
-d CDF

The result will be displayed.

Using the provided pretrained model, our cross-dataset results are reproduced as follows:

Training Data	CDF	DFD	DFDC	DFDCP	FFIW
FF-raw	94.68%	99.12%	80.81%	85.39%	85.55%

Training

1. Download FF++ real videos and place them in ./data/ folder:

.
└── data
    └── FaceForensics++
        ├── original_sequences
        │   └── youtube
        │       └── raw
        │           └── videos
        │               └── *.mp4
        ├── train.json
        ├── val.json
        └── test.json

2. Download landmark detector (shape_predictor_81_face_landmarks.dat) from here and place it in ./src/preprocess/ folder.

3. Run the two codes to extractvideo frames, landmarks, and bounding boxes:

python3 src/preprocess/crop_dlib_ff.py -d Original
CUDA_VISIBLE_DEVICES=* python3 src/preprocess/crop_retina_ff.py -d Original

4. You can download code for landmark augmentation:

mkdir src/utils/library
git clone https://github.com/AlgoHunt/Face-Xray.git src/utils/library

5. Run the training:

CUDA_VISIBLE_DEVICES=* python3 src/train_gfade.py \
src/configs/gfade/base.json \
-n gfade

Top five checkpoints will be saved in ./output/ folder.