This repository contains the code and datasets described in following paper: It’s All in the Touch: Authenticating Users with HOST Gestures on Multi-Touch Screen Devices.
We are actively communicating with volunteers to solicit their consent for anonymous data disclosure, as this clause was not included in the initial agreement.
The code contains three parts: an Android App to collect the sensor data when the screen is touched, a Python Script to extract time- and frequecy- features of the time-series raw data and a Python Script to apply Random Forest and SVM algorithm to build classfiers.
We are about to update the code in few weeks. The Android app has been updated.