Impact of EEG Frequency Bands and Data Separation on the Performance of Person Verification Employing Neural Networks
The paper is devoted to the study of EEG-based people verification. Analyzed solutions employed shallow artificial neural networks using spectral EEG features as input representation. We investigated the impact of the features derived from different frequency bands and their combination on verificat...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-07-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/22/15/5529 |
| _version_ | 1797440783444344832 |
|---|---|
| author | Renata Plucińska Konrad Jędrzejewski Marek Waligóra Urszula Malinowska Jacek Rogala |
| author_facet | Renata Plucińska Konrad Jędrzejewski Marek Waligóra Urszula Malinowska Jacek Rogala |
| author_sort | Renata Plucińska |
| collection | DOAJ |
| description | The paper is devoted to the study of EEG-based people verification. Analyzed solutions employed shallow artificial neural networks using spectral EEG features as input representation. We investigated the impact of the features derived from different frequency bands and their combination on verification results. Moreover, we studied the influence of a number of hidden neurons in a neural network. The datasets used in the analysis consisted of signals recorded during resting state from 29 healthy adult participants performed on different days, 20 EEG sessions for each of the participants. We presented two different scenarios of training and testing processes. In the first scenario, we used different parts of each recording session to create the training and testing datasets, and in the second one, training and testing datasets originated from different recording sessions. Among single frequency bands, the best outcomes were obtained for the beta frequency band (mean accuracy of 91 and 89% for the first and second scenarios, respectively). Adding the spectral features from more frequency bands to the beta band features improved results (95.7 and 93.1%). The findings showed that there is not enough evidence that the results are different between networks using different numbers of hidden neurons. Additionally, we included results for the attack of 23 external impostors whose recordings were not used earlier in training or testing the neural network in both scenarios. Another significant finding of our study shows worse sensitivity results in the second scenario. This outcome indicates that most of the studies presenting verification or identification results based on the first scenario (dominating in the current literature) are overestimated when it comes to practical applications. |
| first_indexed | 2024-03-09T12:13:24Z |
| format | Article |
| id | doaj.art-f1341837effb49e59776fafb56cf17a9 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T12:13:24Z |
| publishDate | 2022-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-f1341837effb49e59776fafb56cf17a92023-11-30T22:50:21ZengMDPI AGSensors1424-82202022-07-012215552910.3390/s22155529Impact of EEG Frequency Bands and Data Separation on the Performance of Person Verification Employing Neural NetworksRenata Plucińska0Konrad Jędrzejewski1Marek Waligóra2Urszula Malinowska3Jacek Rogala4Institute of Electronic Systems, Faculty of Electronics and Information Technology, Warsaw University of Technology, 00-665 Warsaw, PolandInstitute of Electronic Systems, Faculty of Electronics and Information Technology, Warsaw University of Technology, 00-665 Warsaw, PolandLaboratory of Neuroinformatics, Nencki Institute of Experimental Biology, 02-093 Warsaw, PolandLaboratory of Neuroinformatics, Nencki Institute of Experimental Biology, 02-093 Warsaw, PolandInstitute of Physiology and Pathology of Hearing, Bioimaging Research Center, World Hearing Center, Kajetany, 05-830 Nadarzyn, PolandThe paper is devoted to the study of EEG-based people verification. Analyzed solutions employed shallow artificial neural networks using spectral EEG features as input representation. We investigated the impact of the features derived from different frequency bands and their combination on verification results. Moreover, we studied the influence of a number of hidden neurons in a neural network. The datasets used in the analysis consisted of signals recorded during resting state from 29 healthy adult participants performed on different days, 20 EEG sessions for each of the participants. We presented two different scenarios of training and testing processes. In the first scenario, we used different parts of each recording session to create the training and testing datasets, and in the second one, training and testing datasets originated from different recording sessions. Among single frequency bands, the best outcomes were obtained for the beta frequency band (mean accuracy of 91 and 89% for the first and second scenarios, respectively). Adding the spectral features from more frequency bands to the beta band features improved results (95.7 and 93.1%). The findings showed that there is not enough evidence that the results are different between networks using different numbers of hidden neurons. Additionally, we included results for the attack of 23 external impostors whose recordings were not used earlier in training or testing the neural network in both scenarios. Another significant finding of our study shows worse sensitivity results in the second scenario. This outcome indicates that most of the studies presenting verification or identification results based on the first scenario (dominating in the current literature) are overestimated when it comes to practical applications.https://www.mdpi.com/1424-8220/22/15/5529EEGelectroencephalographybiometryverificationneural network |
| spellingShingle | Renata Plucińska Konrad Jędrzejewski Marek Waligóra Urszula Malinowska Jacek Rogala Impact of EEG Frequency Bands and Data Separation on the Performance of Person Verification Employing Neural Networks Sensors EEG electroencephalography biometry verification neural network |
| title | Impact of EEG Frequency Bands and Data Separation on the Performance of Person Verification Employing Neural Networks |
| title_full | Impact of EEG Frequency Bands and Data Separation on the Performance of Person Verification Employing Neural Networks |
| title_fullStr | Impact of EEG Frequency Bands and Data Separation on the Performance of Person Verification Employing Neural Networks |
| title_full_unstemmed | Impact of EEG Frequency Bands and Data Separation on the Performance of Person Verification Employing Neural Networks |
| title_short | Impact of EEG Frequency Bands and Data Separation on the Performance of Person Verification Employing Neural Networks |
| title_sort | impact of eeg frequency bands and data separation on the performance of person verification employing neural networks |
| topic | EEG electroencephalography biometry verification neural network |
| url | https://www.mdpi.com/1424-8220/22/15/5529 |
| work_keys_str_mv | AT renataplucinska impactofeegfrequencybandsanddataseparationontheperformanceofpersonverificationemployingneuralnetworks AT konradjedrzejewski impactofeegfrequencybandsanddataseparationontheperformanceofpersonverificationemployingneuralnetworks AT marekwaligora impactofeegfrequencybandsanddataseparationontheperformanceofpersonverificationemployingneuralnetworks AT urszulamalinowska impactofeegfrequencybandsanddataseparationontheperformanceofpersonverificationemployingneuralnetworks AT jacekrogala impactofeegfrequencybandsanddataseparationontheperformanceofpersonverificationemployingneuralnetworks |