Your signature is your password: Haptic Passwords on Mobile Devices

TL;DR

This paper introduces a novel haptic password system utilizing force-sensitive screens on mobile devices, employing wavelet-based features and adaptive templates to enhance security and accuracy in user authentication.

Contribution

It presents a new behavioral biometric authentication method based on haptic interaction, with a customized classifier and adaptive scheme for improved robustness.

Findings

High authentication accuracy achieved

Robust against forgery attacks

Effective across multiple user postures

Abstract

In this paper, a new behavioral biometric password based on haptic interaction is described. This "haptic password" system is built on a device that has a force-sensitive screen. The user authentication process leverages maximal overlap discrete wavelet transform (MODWT) based feature extraction along with a customized classifier as well as an adaptive template scheme to accommodate users' variations over time. An experimental study was conducted with 29 subjects and we investigated the authentication performance with multiple user postures under simulated forgery attacks. The performance of the model is evaluated based on authentication accuracies. The results suggest that this proposed haptic password system achieves robust high authentication accuracy and scalability, as well as forgery-proof performance.