A test to determine if the detected face is a living person or an impostor trying to defeat the system with a stolen image.

A dynamic algorithm generally involves asking the user to complete a task. For example, asking the user to turn their head to the left or the right or to open their mouth. These movements can be measured using ROC_PITCHYAW. These algorithms generally have low error rates, at the cost of requiring a few seconds of interaction from the user. In order to leverage these methods, permission from Amazon and Google may need to be obtained as they own patents in this field.

A static algorithm attempts to detect the impostor from a single image. For example, detecting if the face image is on a printed piece of paper or computer screen. These algorithms generally have higher error rates, with the benefit that they do not require interaction from the user. Rank One's approach uses a patent pending method based on micro-texture analysis of the user's image. This method requires an image in its original form from the camera, without changes introduced by compression, resizing or other image processing transformations. Use ROC_SPOOF for a static liveness metric.