Step 1: Extract frame from video
python src/generate.py
Step 2: Crop image to scale
python src/generate_patches.py
├───scale_1.0
│ └───liveness_face
│ └───datasets
│ └───datasets_train
| └───label_list.txt
├───scale_2.7
│ └───liveness_face
│ └───datasets
│ └───datasets_train
| └───label_list.txt
└───scale_4
└───liveness_face
└───datasets
└───datasets_train
└───label_list.txt
Config params: ./src/default_config.py
Step 3: training model
python train.py
Step 4: Demo
python src/demo.py --model_dir ./resources/ckpt_onnx/ --input_data ./data/0.mp4
- https://github.com/minivision-ai/Silent-Face-Anti-Spoofing/blob/master/README_EN.md
- https://github.com/Elroborn/Face-anti-spoofing-based-on-color-texture-analysis
- https://github.com/mnikitin/Learn-Convolutional-Neural-Network-for-Face-Anti-Spoofing