The official code for paper "Can We Leave Deepfake Data Behind in Training Deepfake Detector" (NIPS2024 poster)
ProDet is implemented within the framework of DeepfakeBench. The provided code should be placed in the corresponding folders in DeepfakeBench, and test/train on DeepfakeBench as well. There is no additional package required beyond DeepfakeBench and this repository, hence you should easily reproduce our paper with an established DeepfakeBench environment as:
python training/train.py --detector_path ./training/config/detector/prodet.yaml --train_dataset "FaceForensics++" --test_dataset "FaceForensics++" "Celeb-DF-v2" "DFDCP"
Completely organized code and instructions will be made available soon.