A personal touch: recognizing users based on touch screen behavior

  • Authors:
  • Sarah Martina Kolly;Roger Wattenhofer;Samuel Welten

  • Affiliations:
  • Computer Engineering and Networks Laboratory, ETH Zurich, Switzerland;Computer Engineering and Networks Laboratory, ETH Zurich, Switzerland;Computer Engineering and Networks Laboratory, ETH Zurich, Switzerland

  • Venue:
  • Proceedings of the Third International Workshop on Sensing Applications on Mobile Phones
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

The spread of touch screen based smart phones has been constantly increasing over the last few years. However, there are still many open research questions concerning the basic input properties of these devices. We performed a large scale study to research the users' touch screen behavior on standard UI elements. To do so we programmed and published a quiz game that logs touch data and sends it back for evaluation purposes. Over 14,000 persons have played this game so far and sent back statistical data. We use the collected data to present basic touch properties, such as mean hold time and pressure dynamics, and to show that touch screen based input is individual for each person and that one can identify a specific user in a set of 5 users with a precision of about 80%, based on just a few touch events.