A person is someone who will need to be recognized in the future. These are the people you store in our database.
A group is a collection of persons. In order to perform facial recognition, you must define a group first and add persons to it.
Enrollment is the process of uploading images of a person’s face to our database This is how our system will be able to recognize them in the future.
Verification is the process of checking the similarity between two faces. For instance, someone claims to be John Smith. The camera takes an image of them and it is uploaded to our servers alongside John Smith’s ID. Our API will return a True/False decision depending on whether the stored images of John Smith match with the image which was just submitted.
Identification is short for face searching. That is, given an image of person’s face, our API will search the group and it will let you know whether that face belongs to a person in the group.
False Acceptance/Rejection Rate
A false acceptance rate of 1% means that one out of 100 unauthorized people will be granted access. On the other hand, a false rejection rate of 0.5% means that five out of 1,000 authorized people will be falsely rejected by the system. These two errors are inversely proportional and they can never be exactly zero.
These are parameters which allow our system to make a decision. Zenus’ facial recognition API is unique in that it can automatically compute these values for you when you give it the false rejection rate or the false acceptance rate of your choosing.
The amount of time you have to use the identification and verification functionality. You are free to choose when you want to start and for how long the system will be active.