Future biometric security systems will be based on neural networks

A biometrics based security system that is able to learn and capable of cognitive processes is now within reach. Researchers at the University of Calgary’s Biometric Technologies Laboratory are in the midst of finishing a system, which can combine several kinds of biometric information such as fingerprints, facial features, and even iris patterns and use these to correctly identify a subject. The system is also capable of processing information gathered from different sources in order to reach a specific decision.

The biometric security system is able to perform cognitive processes that allow it to learn and evolve. It essentially studies both biometric measurement and behavioral signatures to correctly identify a person. Using both the behavioral and measurable biometrics in identity verification ensures that the results are more accurate and reduces the chance of hitting a false positive or negative. The security system that adapts this kind of technology is not only efficient, but also highly flexible. It is able to process any kind of biometric data along with matching it with a signature, a gait or even postures made by a person of interest.

A system with a neural network such as this can prove to be useful when searching for a suspect in a large crowd. Suspects are constantly changing their plan of attack and the security system designed to seek them out should do the same. This kind of neural network will allow the system to combine key biometric features that can be taken from any database, along with adapting to the environment where the subject is currently located.

For example, a partial facial scan of a suspect standing in the middle of a crowded airport can be used by the system to search for the suspect’s identity. The usual issues such as movement of the crowd, lighting in the airport terminal and cumbersome positioning of security cameras will no longer detract from identification as the system will easily work around it.

Is artificial intelligence the key to a more accurate biometric-based security system?

Article Topics

behavioral biometrics  |  biometrics  |  facial features  |  fingerprint biometrics  |  iris patterns