Suprema Inc(hereafter Suprema), certifies that Suprema FaceStation 2 and FaceLite extract the face data points from the raw images and create a template from this data. The face template is not a raw image. It is created by using Suprema’s patented NIR technology and algorithm, and the face template is a dataset that is a fusion of 2D and 3D features of a user's face which is acquired by a visual camera and NIR camera.
In order to secure this template, cryptographic tools and encryption methods are used such as 256bit AES encryption making it very difficult to access the necessary data to reverse engineer the process. The ability to reduplicate the original image and prove the identity via a face recognition expert is not possible.
FaceStation 2 and FaceLite also calculate a matching score when executing an authentication. The matching score is a calculation of the match between a pre-registered face template and the actual face to be authenticated.
The matching score helps protect FaceStation 2 and FaceLite against spoofing.
Suprema also certifies that FAR(False Acceptance Rate) and FRR(False Rejection Rate) in face recognition algorithm in FaceStation 2 and FaceLite are as below.
▶ FRR: Less than or equal to 1%
▶ FAR: Less than or equal to 0.00002% (1/5,000,000)