Face Spoofing Detection using Depth Wise Convolution

Pytorch implementation of Depth Wise Convolution to perform the task of Face Spoofing Detection. The code implementation was done following the paper [A Face Spoofing Detection Method Based on Domain Adaptation and Lossless Size Adaptation] (https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9056475). The Domain Adaptation Layer (partial) and Lossless Size Adaptation methodologies were avoided in this implementation.

Getting Started

Installing

• Install PyTorch and dependencies from https://pytorch.org.

Training

• Download any dataset according to your choice.
• Go to config.py and make necessary changes according to the helping texts given.
• Run the following command:

python run.py

Inference

• Go to the liveness_testing folder
• Copy and paste the desired trained model in this directory
• Go to liveness_demo.py and change the model path
• Run the following command:

python liveness_demo.py