On the Optimality of the Simple Bayesian Classifier under Zero-One Loss
Machine Learning - Special issue on learning with probabilistic representations
Pattern Recognition Letters - In memory of Professor E.S. Gelsema
Person Authentication Using Brainwaves (EEG) and Maximum A Posteriori Model Adaptation
IEEE Transactions on Pattern Analysis and Machine Intelligence
Unobtrusive biometric system based on electroencephalogram analysis
EURASIP Journal on Advances in Signal Processing
Iris recognition: an emerging biometric technology
ISPRA'07 Proceedings of the 6th WSEAS International Conference on Signal Processing, Robotics and Automation
Handbook of Fingerprint Recognition
Handbook of Fingerprint Recognition
Personal verification using palmprint and hand geometry biometric
AVBPA'03 Proceedings of the 4th international conference on Audio- and video-based biometric person authentication
De-noising by soft-thresholding
IEEE Transactions on Information Theory
| Hi-index | 0.00 |
Identification of individuals is ubiquitous with increasing reliance by financial and governmental organizations on reliable and robust personal recognition systems to determine and confirm the identity and policy constraints for specific individuals in 'real-time' when reacting to service requests. The traditional identification approaches to user validation are not robust and deficiencies in such approaches are becoming increasingly apparent in the current information-oriented society. With developments in research into the human brain, biometric methods based on brain wave signals have received increased attention as an effective approach to user validation as an individuals' brain wave signals cannot be duplicated, discarded or stolen. Targeted at pervasive systems and the identified deficiencies in traditional approaches to user identification and validation, an electroencephalogram (EEG)-based biometric system for use in pervasive environments is proposed in this paper. A significant problem of EEG-based biometrics in pervasive environment is the requirement of real-time and convenience. In our study, only one active electrode with a portable EEG collection device was used and no other instructions to users for convenience; in addition, the signal analysis methods we used were efficient to achieve less time consumption. In our prototype system, 11 subjects were identified with recognition accuracy in the range 66.02% to 100%; the recognition accuracy increased with increases in the EEG sample time; and the computational time of signal analysis was about 0.5s. The low computational time of EEG data analysis validates this model when implemented in pervasive environments. For differing applications we can define a suitable balance point to optimize the conflicting demands of data collection time length and recognition accuracy.