Running Presentation Attack Detection Experiments

The bob.pad packages provide open source tools to run comparable and reproducible presentation attack detection (PAD) experiments. The API of running experiments is designed on top of sklearn’s API to be as simple and familiar as possible. On top of that, we use Bob Pipelines’s sample-based API to extend the sklearn’s API to work with sample-based data. Make sure you have reviewed the documentation of Bob Pipelines before continuing here.

This package is the base package of the PAD framework. It provides the core functionality to run PAD experiments and to evaluate the results. However, the implementation of the PAD algorithms are provided in modality specific packages such as:

  • bob.pad.face

  • bob.pad.voice (at the moment of writing, this package is not maintained anymore and does not work with the new version of bob.pad.base.)

Follow the documentation below to learn how to use the PAD framework.

Users Guide

Reference Manual

Indices and tables