Loading...

EEG based biometrics using emotional stimulation data

Khalil, R ; Sharif University of Technology | 2018

803 Viewed
  1. Type of Document: Article
  2. DOI: 10.1109/R10-HTC.2017.8288949
  3. Publisher: Institute of Electrical and Electronics Engineers Inc , 2018
  4. Abstract:
  5. EEG based biometrics using linear Support Vector Machine (SVM) is proposed in this paper. Human identification using electroencephalographic signal was done in this research. Reliability of most of the biometrics systems is not up to the mark because of the possibility of being faked or duplicated. Here, the brain signatures were used as a possible biometric identifier. A Database for Emotion Analysis using Physiological Signals containing 40 trials from each participant was used. International 10-20 system of EEG electrode placement was employed and data from Cz electrode was taken for this research. Some researches showed nice performance with few subjects. Here, 20 subjects were used from the dataset for the system. With this, the system gives 77% mean precision and at the same time 100% detection accuracy. © 2017 IEEE
  6. Keywords:
  7. Biometrics ; Electroencephalogram (EEG) ; Power spectral density (PSD) ; Support vector machine (SVM) ; Biometrics ; Electrodes ; Spectral density ; Support vector machines ; Biometric identifiers ; Electro-encephalogram (EEG) ; Electroencephalographic signals ; Emotional stimulations ; Fast fourier transformation (FFT) ; Linear Support Vector Machines ; Physiological signals ; Power spectral densities (PSD) ; Electroencephalography
  8. Source: 5th IEEE Region 10 Humanitarian Technology Conference, R10-HTC 2017, 21 December 2017 through 23 December 2017 ; Volume 2018-January , February , 2018 , Pages 246-249 ; 9781538621752 (ISBN)
  9. URL: https://ieeexplore.ieee.org/document/8288949