The general premise will be to clone our Chapter 8, Creating a Facial Expression Platform on IBM Cloud, project and create a class or classes to be used as a simple biometric signature for myself; one loaded with a dozen images of, well, my face, as well as a negative class loaded with a dozen images of random faces (none will be of me).

In addition, there will be at least one other class, designed to be a biometric signature of another individual. So we expect that, in this chapter's project, we will have the following three classes:

• Jim class
• Individual two or other classes
• Negative class

The final step will be to use a mobile device to take several photos of myself and others and, without preparation or any processing of the images, submit the facial images to our project and record the results: matched or not matched. These image submissions will be considered our attempts for biometric authentication.

Perhaps to test the project fairly, we will need to be sure to submit several images of myself (none of which would be part of the model's class definitions my biometric signature) and several faces that are not mine and record the results of each submission.

Finally, we will evaluate and record the results.