Fourier Synchrosqueezing Transform-ICA-EMD Framework Based EOG-Biometric Sustainable and Continuous Authentication...

Area of Science:

• Biometric Authentication
• Signal Processing
• Human-Computer Interaction

Background:

• Limited research exists on Electrooculography (EOG)-based biometrics using eye movements or blinks.
• EOG signals possess unique and permanent traits suitable for individual identification.
• The increasing need for secure online authentication in e-learning, e-banking, and e-government systems is critical.

Purpose of the Study:

• To investigate the performance of a robust FSST-ICA-EMD framework for EOG-biometric authentication.
• To evaluate the system's effectiveness using voluntary eye blink movements.
• To present a sustainable, continuous, and reliable EOG-biometric authentication method for digital applications.

Main Methods:

• Utilized a robust framework combining Fourier Synchrosqueezing Transform (FSST), Independent Component Analysis (ICA), and Empirical Mode Decomposition (EMD).
• FSST was employed for accurate temporal-spatial property extraction from EOG signals.
• ICA was used for decomposing multi-channel EOG recordings, and EMD isolated EOG signals from background EEG.

Main Results:

• Achieved high recognition accuracy (ACC: ≥99.99%) and Area Under the Curve (AUC: 0.99) in one-versus-others verification.
• Demonstrated high True Acceptance Rate (TAR: ≥98%) and low False Acceptance Rate (FAR: ≤3.33%) across seven individuals.
• This represents the first exploration of the FSST-ICA-EMD framework for EOG-biometric authentication using voluntary eye blinks.

Conclusions:

• The proposed FSST-ICA-EMD framework offers a highly accurate and robust EOG-biometric authentication system.
• Voluntary eye blink movements can be effectively utilized for reliable user authentication.
• This approach provides a scalable and affordable solution for continuous authentication in digital environments, including e-learning.