Classifying accelerometer data via hidden Markov models to authenticate people by the way they walk
| Autor: | Claudia Nickel, Christoph Busch |
|---|---|
| Rok vydání: | 2013 |
| Předmět: |
Password
Authentication Engineering User Friendly Biometrics Computer science business.industry Feature extraction Word error rate Aerospace Engineering Pattern recognition Accelerometer Data set Gait (human) Space and Planetary Science Human–computer interaction Embedded system Artificial intelligence Electrical and Electronic Engineering Hidden Markov model business Mobile device Test data |
| Zdroj: | ICCST |
| ISSN: | 0885-8985 |
| DOI: | 10.1109/maes.2013.6642829 |
| Popis: | As owners of mobile devices tend to deactivate their security settings, data on these devices is often insufficiently protected [1]. One reason for this is that most mobile devices do only offer the authentication via PIN or password, which requires explicit interaction and thus is time consuming and not very user friendly. To solve this problem, an alternative unobtrusive authentication method based on gait is proposed in this article. There are two main advantages of this approach. First, gait can be captured via acceleration sensors, which are already integrated into most smart phones. Hence, there are no additional hardware costs for deploying this method. Second, gait recognition does not require explicit user interaction during verification as the phone does it literally "on the go." These two factors make accelerometer-based biometric gait recognition a very user friendly method, which does not require extra interaction time. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|