Active Face Liveness Detection Using Mediapipe

Active liveliness detection is one of the important aspects of biometric authentication that helps us detect whether a person is alive and physically present in authentication process. In this paper, we are going to discuss all the details of active liveliness detection system as well as passive liveliness system, and we will make our own active liveliness detection system using Mediapipe library of python. Active liveliness detection system requires the users to perform specific actions such as moving their head left, right, up and down, smiling and blinking. If the user successfully performs the above actions, we will ensure they are present during the authentication process. We will use Mediapipe library to detect facial landmarks and specific patterns to find head movements and facial expressions. We will use Eye Aspect Ratio (EAR) to detect blinking action. To measure the distance between landmark points, we will use Euclidean distance formula. In the same way, we will find happiness using the Happiness Score formula with the specific landmark points.