Types Used for Voice Interpretation

Now that we know what constitutes the voice biometric and how it can be captured, we need to know what types of algorithms are used. The algorithms used to match and enroll the voice biometric fall into the following general categories:

·         Fixed phrase authentication

·         Fixed vocabulary authentication

·         Flexible vocabulary authentication

·         Text-independent authentication

How Can This Bio metric Be Spoofed?

As discussed in the opportunity of this episode, it was conventional that even humans can be fooled into thinking we distinguish a voice when we do not. If this is the case, it is authentic and to be predictable that a voice biometric system could be fooled as well. While it is generally conventional that voice biometrics do not make available the same level of FAR as other biometrics, voice biometrics offer other very beautiful attributes such as high user reception and low hardware cost. The term “low cost” is relative to the quality of the classification and thus the ability to productively enroll and verify. This is based on the ambient noise in the authentication and enrollment location and the quality demanded by the underlying algorithm. It is ideal to use the same type of enrollment and authentication device. This way, acoustical differences between different enrollment and authentication devices can be minimized.

Voice biometrics is like any other biometric: susceptible to some level of spoofing. Attacks on a voice biometric system fall into the following categories

Voice biometrics presents an attractive option for biometric network security. They offer relative ease of use with also comparatively cheap hardware. The biometric is not invasive to the user and is used on a daily basis. The drawbacks are the possible amount of conscription to have a strong enough lexicon to use and the propensity of the biometric to impact from both environmental and physiological conditions, like catching a cold. As such, to contain for these drawbacks, the choice of FAR versus FRR needs to balance. Again, it is a question of convenience versus security. In my opinion, there is biometrics better suited to network authentication than voice biometrics.