XceptionNet from our paper trained on our FaceForensics++ dataset. Besides the full image models, all models were trained on slightly enlarged face crops with a scale factor of 1.3. The models were trained using the Face2Face face tracker, though the detect_from_models.py file uses the freely available dlib face detector.

Note that we provide the trained models from our paper which have not been fine-tuned for general compressed videos. You can find our used models under this link.

Setup:

• Install required modules via requirement.txt file
• Run detection from a single video file or folder with

python detect_from_video.py
-i <path to input video or folder of videos with extenstion '.mp4' or '.avi'>
-m <path to model file, default is imagenet model
-o <path to output folder, will contain output video(s)

from the classification folder. Enable cuda with --cuda or see parameters with python detect_from_video.py -h.

• python 3.6
• requirements.txt