Related Experiment Video
Updated: Aug 23, 2025
06:34
A Single-Channel and Non-Invasive Wearable Brain-Computer Interface for Industry and Healthcare
Published on: July 7, 2023
2.5K
The single-channel dry electrode SSVEP-based biometric approach: data augmentation techniques against overfitting for
Kutlucan Gorur1, Beyza Eraslan2
1Electrical and Electronics Engineering, Bandirma Onyedi Eylul University, 10200, Balikesir, Turkey. kgorur@bandirma.edu.tr.
Physical and Engineering Sciences in Medicine
|November 1, 2022
Summary
This study introduces a novel, user-friendly biometric system using steady-state visually evoked potentials (SSVEP) and Recurrent Neural Networks (RNNs). The system achieves high accuracy for individual identification, offering a low-cost authentication solution.
Area of Science:
- Biometrics
- Neuroscience
- Signal Processing
Background:
- Traditional electroencephalography (EEG) biometrics face challenges like noise and effort.
- Steady-state visually evoked potential (SSVEP) biometrics offer a high signal-to-noise ratio and require no user training.
Purpose of the Study:
- To compare multi-channel SSVEP biometrics with a novel single-channel, single-trial SSVEP approach using Recurrent Neural Networks (RNNs).
- To evaluate the efficacy of data augmentation strategies with RNNs for SSVEP-based biometrics to prevent overfitting.
Main Methods:
- Implementation of a single-channel, single-trial SSVEP biometric system utilizing dry electrodes.
- Application of Recurrent Neural Networks (RNNs) with data augmentation for biometric recognition.
- Performance evaluation using accuracy, sensitivity, specificity, and F-scores.
Main Results:
- The single-channel SSVEP-based biometric system achieved up to 100% accuracy.
- Sensitivity and specificity scores exceeded 97%, with F-scores also above 97% for 11 subjects.
- The RNN deep models demonstrated promising results for individual identification.
Conclusions:
- The proposed single-channel SSVEP biometric approach using RNNs is a low-cost, user-friendly, and reliable method for individual identification.
- This technique has potential for significant applications in authentication and security systems.
