A novel nonintrusive user authentication method based on touch gestures for smartphones

Chien Cheng Lin, Chin Chun Chang, Deron Liang

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

Novel nonintrusive authentication mechanisms for smartphones have recently been investigated to complement existing protection mechanisms. A major reason is that a nonintrusive approach can continually authenticate users without interrupting their operations. In this paper, a novel approach to nonintrusive authentication of smartphone users involving flick gestures is proposed. Sixteen histogram-based features of flick gestures, including five novel features related to the habitual touch positions of a user, were used for authentication. Empirical results based on 51 participants indicated that the proposed approach was feasible. The equal error rate of the proposed approach was approximately 4.37% ∼ 5.13% when the histogram-based features were calculated based on 30 flicks and decreased to approximately 2.35% ∼ 2.99% for features based on 60 flicks. In addition, the experimental results revealed that the five novel features proposed are effective. Possible applications and limitations of the proposed approach are discussed.

Original languageEnglish
Pages (from-to)801-810
Number of pages10
JournalJournal of Internet Technology
Volume16
Issue number5
DOIs
StatePublished - 2015

Keywords

  • Histogram-based features
  • Nonintrusive authentication
  • Smartphone
  • Touch screen sensor

Fingerprint

Dive into the research topics of 'A novel nonintrusive user authentication method based on touch gestures for smartphones'. Together they form a unique fingerprint.

Cite this