Gender impact on the identification based on EEG

Jinghai Yin, Zhendong Mu

Institute of Information Technology, Jiangxi University of Technology, Nanchang 330098, China

To study the genders impact on identification, this paper analysis the electroencephalograph (EEG) of eight male subjects and seven female subjects. In order to reduce the noise signal interference, the high pass and low pass were used to cut extra frequencies, and in order to prominent the feature signal, the power spectrum method was used to convent the time domain signal to frequency domain, and then fisher distance was used to extraction the feature. All EEG signal was acquired by neurescan, and the EEG signal was evoked by VEP method used subjects photo. The experiment was divided into three models: all subjects were the same sex, added some opposite sex, added some stranger. The analysis results show, to model 1, the correct recognition rate for male subjects, average is 88.50, and this for female, average is 92.51%; the false recognition rate for male subjects, average is 30.84%, and this for female, average is 27.67%,this result indicates VEP can be used as identification tool, the results of model 2 and model 3 show weather opposite sex or stranger should affect the correct recognition rate, but to male subject, the opposite sex effect is greater than stranger, to female, the result were reversed. The results also show noise photo affected female lower than male.