This code package is related to the paper:
M. Nerini, E. Favarelli, and M. Chiani, "Augmented PIN Authentication through Behavioral Biometrics," Sensors, 2022.
The dataset folder contains 12 datasets obtained by 12 different students. Each dataset is a .txt file representing a 470 x 17 matrix in which:
- each row is a sampled digit.
- each column is a feature: the first is the pressed digit and the following are motion sensor values.
The PIN_authentication.ipynb Jupiter Notebook contains the code to replicate the results in the paper.
The files rocMat.mat and times.mat have been obtained throguh PIN_authentication.ipynb and are attached for an easier replication of the figures.