Skip to content

Learning Generalized Spoof Cues for FaceAnti-spoofing

License

Notifications You must be signed in to change notification settings

yangjian1218/LGSC-for-FAS

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

31 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

LGSC-for-FAS

This repository contains code of "Learning Generalized Spoof Cues for FaceAnti-spoofing (LGSC)", which reformulate face anti-spoofing (FAS) as an anomaly detection problem and learn the discriminative spoof cues via a residual-learning framework.


Framework

LGSC-Framework

FaceForensics Benchmark

Based on the LGSC, we achieved 1st Place on FaceForensics benchmark FaceforensicsBenchmark

Dataset and Preprocessing

We use two different compression rates: c23(medium compression) and c40(high compression) videos in FaceForensics++ dataset as our training data. As a preprocessing step we extract crop faces frame from raw videos and totally we get 0.96 million face frames, which contrains 0.16 million pristine frames and 0.8 million manipulated frames. Further, we balanced the proportion of positive and negative samples during training.

Technology Detail

Our deep network is implemented on the platform of PaddlePaddle(dygraph).

  • Download and preproces the dataset
FaceForensics/data
├── manipulated_sequences
│   └── DeepFakeDetection
│   └── Deepfakes
│   └── Face2Face
│   └── FaceSwap
│   └── NeuralTextures
├── original_sequences
│   └── actors
│   └── youtube
│       └── c23
│       └── c40
│           └── images
├── train_add_train.txt
├── train_val_train.txt
└── faceforensics_benchmark.txt
python convert.py

./pretrained
├── resnet18-5c106cde.pth
└── resnet18-torch.pdparams
  • Train
python train.py
  • Test
python test.py

Citation

@inproceedings{
    feng2020deep,
    title={Learning Generalized Spoof Cues for Face Anti-spoofing},
    author={Haocheng Feng and Zhibin Hong and Haixiao Yue and Yang Chen and Keyao Wang and 
    Junyu Han and Jingtuo Liu and Errui Ding},
    year={2020},
    eprint={2005.03922},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}

About

Learning Generalized Spoof Cues for FaceAnti-spoofing

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Languages

  • Python 100.0%