Face Liveness Detection
Verify the human presence or inanimate objects. IdentifAI Face Liveness Fraud Detection System is a machine learning technology to give recommendation accept real human presence and reject spoof objects, especially inanimate objects that resembles a face such as replay video, mask, or printed photos to minimize fraud risks.
Detecting Spoof Component
Examine the type of spoof artifact such as image from printed photo, replay video and mask.
Capable for Detecting
Low Quality Image
Low-quality image detection: Review the image quality input from the selfie to minimize rejection error
Gesture Identification
Gesture Identification (Checking the human presence by following the instruction and analyzing face landmark for each gesture)
Active & Passive Combination (SDK Deployment)
Hold for 3 seconds to capture the image
Verifying image
Image Verification
Liveness
result sent to
the user
Cloud API performs passive liveness check
On the SDK mobile, it performs active liveness check
Checking activities: The user needs to follow 3 pose instructions within 10s. If this step takes more than 10s, it will return to the beginning, which will ask the user to verify their face detection. If the system finds two different faces on a sequence, the process will be repeated.
Passive Liveness check process:
Spoof component identification
Verify parametric liveness score
Try our Nodeflux Face Liveness demo apps
Passive Liveness (API Deployment)
Image sent to
Nodeflux passive liveness API
Liveness
result sent to the user
Image Verification
Prevent Fraud Potential
Our Face Liveness detection can capture the spoofed image, and ensure the presence of real person who do registration. The system classify the likely spoof object with obvious characteristic to be rejected.
