Straightaway, we need to have a sample of 10 images of my face. The images should be of reasonable quality and, for best results, close-up photos taken over a period of time so that the algorithm can have some accounting for variances in my appearance.
Here are the images that will be used to create my biometric signature:
Next, we need to establish a biometric signature (with the same size sample) for another individual:
Finally, the negative class of 10 random faces (of course, borrowed from our project) needs to be set:
A reminder from Watson docs: The create a classifier call requires that you provide at least two example ZIP files: two positive examples files or one positive and one negative file. The negative examples defines what the updated classifier is not. It is not used to create a class within the created classifier. Negative example files will contain images that do not match the subject of any of the positive examples. In a single call, there can be only one single example specified.