Face PAD for Silicone mask-based attack detection

This package is part of the Bob toolkit and it allows to reproduce the experimental results in the following paper:

@article{XCSMAD-TBIOM-2019,
    title = {{Multispectral Deep Embeddings As a Countermeasure To Custom Silicone Mask Presentation Attacks}},
    author = {{K. Kotwal, S. Bhattacharjee, and S. Marcel}},
    journal = {{**IEEE Transactions on Biometrics, Behavior, and Identity Science**}},
    volume = {00},
    number = {0},
    pages = {000--000},
    month = {{0}},
    year = {2019},
}

If you use this package and/or its results, please cite the paper [TBIOM2019].

User guide

• Pre-requisites and Setting-up the experiments
  • Downloading the dataset
  • Downloading the face recognition CNN model
  • Setting up annotations directory
  • Generating set of commands
• Executing Face PAD experiments
  • A. Single Channel PAD
  • B. Feature Fusion
  • C. Score Fusion
  • D. Cross Validation
• Executing Vulnerability Analysis Experiment
• References