Related Experiment Video
Updated: Oct 16, 2025
05:48
Author Spotlight: Investigating the Impact of Emotional Prosodies on Voice Recognition and Perception
Published on: August 9, 2024
1.7K
Chronic Study on Brainwave Authentication in a Real-Life Setting: An LSTM-Based Bagging Approach
Liuyin Yang1, Arno Libert1, Marc M Van Hulle1
1Department of Neuropsychology and Physiology, KU Leuven, 3000 Leuven, Belgium.
Biosensors
|October 22, 2021
Summary
This study demonstrates the feasibility of electroencephalography (EEG) brainwave authentication outside the lab. Using a commercial headset and advanced AI, it achieves over 92% accuracy for secure data access.
Area of Science:
- Neuroscience
- Biometrics
- Cybersecurity
Background:
- Increasing need for secure data access in the digital age.
- Exploration of biometric authentication methods beyond traditional techniques.
- Emerging interest in electroencephalography (EEG) for brainwave-based security.
Purpose of the Study:
- To validate EEG-based authentication in real-world, out-of-laboratory conditions.
- To assess the efficacy of using commercial, dry-electrode EEG headsets for authentication.
- To evaluate a novel deep learning approach for EEG signal decoding.
Main Methods:
- Utilized a commercial dry-electrode EEG headset for chronic recordings.
- Employed a Long Short-Term Memory (LSTM)-based network with bootstrap aggregating (bagging).
- Applied a multitask scheme involving performed and imagined motor tasks for decoding EEG signals.
Main Results:
- Achieved authentication accuracy of 92.6% for performed motor tasks.
- Reported 92.5% accuracy for imagined motor tasks and 93.0% for combined tasks.
- Demonstrated improved performance over standard LSTM using the proposed bagging method.
Conclusions:
- EEG-based authentication is feasible in real-world, non-laboratory settings.
- The proposed LSTM with bagging method enhances authentication performance.
- This approach is recommended for time- and data-limited authentication scenarios.
