Smartwatch-based transcription biometrics

Abstract

Handwriting analysis and the identification of people based on individual typing patterns are of great interest in the field of biometrics. The increasing need for user authentication in our modern society has underscored the need for smarter and more automated security protocols. The goal of this research is to explore the feasibility of using relatively inexpensive commercial smartwatches as a reliable platform for transcription-based biometrics. This study demonstrates that it is possible to accurately identify individuals performing transcription tasks (i.e. handwriting sentences, typing on a keyboard, etc.) using only a smartwatch.