Active Liveness requires the user to perform actions to either turn left or right or blink. The activity-based Liveness is spoofed by having the image cut out and turned in the respective direction. Additionally, activity-based Liveness slows down the authentication and provides a bad user experience.
To be considered active, the user must take some action: to blink, turn left, or turn right. The image is clipped out and rotated in the appropriate direction to spoof the activity-based Liveness. Additionally, Liveness based on activity slows down the login and degrades the user experience.
In Passive Liveness, The algorithm does not leave any tips for the user to fake the liveness detection. The various techniques used in Passive Liveness detection are
1. Using an additional thermal sensor :
This approach requires keeping an additional thermal sensor on the hardware, thereby requiring customised hardware
reliability - good
user experience - good
adoption - bad
2. Capturing a short video:
The algorithm processes a video by observing the micro-motions and small movements. The process requires a noticeable amount of time, affecting the user experience.
reliability - good
user experience - bad
adoption - bad
3. Examining a single selfie image:
The algorithm processes a single selfie image for liveness detection. This method is fast and provides the best user experience on a mobile or web application.
reliability - good
user experience - good
adoption - good
4. Flashing various kinds of lights:
The device flashes various light colours on the face in this method. However, this method does not work well in sunlight.
reliability - bad
user experience - bad
adoption - good