BioID is designed to enable the use of several different biometrics. Any of them can be combined in a so-called "fusion" to get one result, which has a better accuracy than using each of the single modalities. This technique of fusing together several biometrics is called multimodal biometrics. (Please note that there is a big difference to so-called "layered" biometrics - their several biometric systems are chained up and used one after the other).
As renowned scientist John Daugman claimed in his article (Combining Multiple Biometrics, John Daugman, The Computer Laboratory, Cambridge University), "A strong biometric is better alone than in combination with a weaker one... when both are operating at their cross-over points." "To reap any benefits from combination", he further explains, "the equations above show that the operating point of the weaker biometric must be shifted...".
In order to reap this benefit and increase the accuracy of the complete system by combining several (distinct) features, a well-designed multimodal system like BioID does not just add several biometrics arbitrarily. It takes into account the different characteristics of each modality and adapts each of them within the sophisticated fusion algorithm. By doing it this way, the resulting system can reach accuracies far beyond each single modality (a sample is given in section About Classification).
For the interested reader, we recommend the outstanding work of Prof. Josef Kittler of Surrey University. Prof. Kittler has been working on multimodal combinations for many years and published a large amount of scientific papers where he investigated many fusion methods. And he could prove both theoretically and in experiments that a good decision rule always increases the classification results. Here is only a small list of articles recommended:
A sample, showing the increase in classification performance by fusing the results of three classifiers, can be found in the section About Classification.