Adaptive biometric systems: data set

PRA group collected the Template Update Fingerprint and Face data set (TUFF) aimed to study the phenomeno of intra-class variations effects, including some medium-term effects due to aging when using a multi-modal system with fingerprint and face biometrics. To this aim, the following features characterize this data set:

  • 100 different people
  • 100 images per subject and per biometric, collected over a time-span of two years
  • face images collected by varying appearance, expressions and lighting conditions (indoor images)
  • fingerprint images collected by varying moisture, pressure, position with respect to the scanner surface
  • two different fingers: right index and thumb

Adopted fingerprint scanner is Biometrika FX2000, whilst the Logitech Quickcam Pro5000 has been adopted for capturing faces.

This data set is available by contacting Dr. Gian Luca Marcialis (marcialisATdieeDOTunicaDOTit).