PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
August 10, 2016IEEE Transactions on Information Forensics and Security99 citations

Continuous Authentication Using One-Dimensional Multi-Resolution Local Binary Patterns (1DMRLBP) in ECG Biometrics

View Full Paper
Ask AI
Bookmark
Share

Population

1,012 single session subjects and 82 multiple sessions subjects from the University of Toronto database

Comparison

One-Dimensional Multi-Resolution Local Binary… vs State-of-the-art work and biometric system based…

Design

Other

Key result

Continuous authentication using 1DMRLBP on ECG signals achieved an equal error rate of 7.89% compared to 12.30% for state-of-the-art methods.

Authors

WLWael LouisMKMajid KomeiliDHDimitrios Hatzinakos

Discussion

Loading...

Member takes

Overview

May enhance ECG biometric security in cardiology systems; leaves open prospective validation before clinical use.

Structured PICO

P
Population
1,012 single session subjects and 82 multiple sessions subjects from the University of Toronto database (UofTDB)
I
Intervention
One-Dimensional Multi-Resolution Local Binary Patterns (1DMRLBP) feature extraction and continuous authentication system
C
Comparator
State-of-the-art work and biometric system based on raw time-samples
O
Outcome
Equal error rate (EER), false rejection rate, and false acceptance rate

The proposed 1DMRLBP feature extraction method for ECG signals improves continuous biometric authentication accuracy compared to existing methods.

Main Result

Absolute Event Rate: 7.89% vs 12.3%

Cite This Study

Louis et al. (2016) studied ECG biometrics (n=1,012). One-Dimensional Multi-Resolution Local Binary Patterns (1DMRLBP) vs. State-of-the-art work and raw time-samples was evaluated on Equal error rate (EER) on single session subjects. Continuous authentication using 1DMRLBP on ECG signals achieved an equal error rate of 7.89% compared to 12.30% for state-of-the-art methods.

synapsesocial.com/papers/6a1d62251e7099f69105585ahttps://doi.org/10.1109/tifs.2016.2599270
View Full Paper
Ask AI
Bookmark
Share