ECG Biometric Recognition by Convolutional Neural Networks with Transfer Learning Using Random Forest Approach

Ankur,; Bharti, Manoranjan Rai

doi:10.1007/978-981-19-7524-0_16

Cited by 2 publications

(1 citation statement)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The cornerstone of robust security is in ensuring that only authorized organizations or individuals can access sensitive data [ 2 ]. Previous studies have proposed various approaches to address these security issues, including biometric authentication and encryption using physiological signals [ 3 , 4 , 5 , 6 ].…”

Section: Introductionmentioning

confidence: 99%

Preliminary Study of Novel Bio-Crypto Key Generation Using Clustering-Based Binarization of ECG Features

Hwang,

Lee,

Kwon

et al. 2024

Sensors

View full text Add to dashboard Cite

In modern society, the popularity of wearable devices has highlighted the need for data security. Bio-crypto keys (bio-keys), especially in the context of wearable devices, are gaining attention as a next-generation security method. Despite the theoretical advantages of bio-keys, implementing such systems poses practical challenges due to their need for flexibility and convenience. Electrocardiograms (ECGs) have emerged as a potential solution to these issues but face hurdles due to intra-individual variability. This study aims to evaluate the possibility of a stable, flexible, and convenient-to-use bio-key using ECGs. We propose an approach that minimizes biosignal variability using normalization, clustering-based binarization, and the fuzzy extractor, enabling the generation of personalized seeds and offering ease of use. The proposed method achieved a maximum entropy of 0.99 and an authentication accuracy of 95%. This study evaluated various parameter combinations for generating effective bio-keys for personal authentication and proposed the optimal combination. Our research holds potential for security technologies applicable to wearable devices and healthcare systems.

show abstract