Dataset Description
Training data:
The training data is presented in: SynthASpoof: Developing Face Presentation Attack Detection Based on Privacy-friendly Synthetic Data [1]. The bona fide samples in SynthASpoof are synthetically generated and the attack samples are collected by presenting such synthetic data to capture systems in a real attack scenario. SynthASpoof [1] contains 25,000 bona fide and 78,800 attack samples.
To acquire the dataset, please visit the repository Github. The more description about dataset, please refer to work SynthASpoof.
[1] Meiling Fang, Marco Huber, and Naser Damer: SynthASpoof: Developing Face Presentation Attack Detection Based on Privacy-friendly Synthetic Data. 2023.
Test data:
To reflect the real-world scenario, the submitted algorithms will be evaluated on authentic face PAD datasets.
Data Pre-processing:
In the competition, the training data should be pre-processed, following the crop face process here.
In the evaluation, the test data follows the same pre-processing procedure.