| Title | Electroencephalogram Based Biometrics: A Fractional Fourier Transform Approach |
| Publication Type | Conference Paper |
| Year of Publication | 2018 |
| Authors | Keshishzadeh, Sarineh, Fallah, Ali, Rashidi, Saeid |
| Conference Name | Proceedings of the 2018 2Nd International Conference on Biometric Engineering and Applications |
| Publisher | ACM |
| ISBN Number | 978-1-4503-6394-5 |
| Keywords | biometrics, channel selection, classification, fractional fourier transform, Metrics, pubcrawl, resilience, Resiliency, Scalability, security, Time Frequency Analysis, verification |
| Abstract | The non-stationary nature of the human Electroencephalogram (EEG) has caused problems in EEG based biometrics. Stationary signals analysis is done simply with Discrete Fourier Transform (DFT), while it is not possible to analyze non-stationary signals with DFT, as it does not have the ability to show the occurrence time of different frequency components. The Fractional Fourier Transform (FrFT), as a generalization of Fourier Transform (FT), has the ability to exhibit the variable frequency nature of non-stationary signals. In this paper, Discrete Fractional Fourier Transform (DFrFT) with different fractional orders is proposed as a novel feature extraction technique for EEG based human verification with different number of channels. The proposed method in its' best performance achieved 0.22% Equal Error Rate (EER) with three EEG channels of 104 subjects. |
| DOI | 10.1145/3230820.3230821 |
| Citation Key | keshishzadeh_electroencephalogram_2018 |
Electroencephalogram Based Biometrics: A Fractional Fourier Transform Approach