IDmission > Facial Recognation
Passive Liveness for Facial Recognation

IDmission is a leader in the use of Passive Liveness for identity verification. With over a decade of development of Machine Learning based facial recognition apps, IDmission’s passive liveness is used worldwide to eliminate identity fraud and use facial biometrics to build identity databases.

Group 196 (1)

Passive Liveness Detection

Our passive liveness detection automatically detects and rejects presentation attacks, including photos and videos, print-outs, masks and deep fakes. IDmission is one of the few companies meeting global ISO 30107-3 Presentation Attack Detection (PAD) Level 2 standards.

Passive Liveness is invisible by design. With just a selfie, a live face can be verified with no need to blink, smile, or turn your head. Masks, glasses and facial hair can all be detected.

See it in Action
Person - Is this a real human being, not a statue, mask, or virtual persona?
Physical Presence - Is this person authenticating in real time, and not using a photo, video, or deep fake?
Precise identity - Is this person precisely who they claim to be, and do they have the rights to access the account or information?
Passive Liveness

Machine Learning Liveness Models

IDmission Machine Learning models are trained on millions of identity records from over 200 countries, creating a globally effective and unbiased platform for building identity systems.

IDmission uses a convolutional neural network (CNN) to solve the problem of detecting liveness from realtime images.

Machine Learning Liveness

The CNN uses deep learning (multiple layers) as follows:
  • Layers - 160 layers deep
  • Trainable Parameters: 3,422,000
  • Performs a 3D convolution using 3 pixel x 3 pixel x 1 pixel filters

    Each day, tens of thousands of new facial verifications are performed, with results from all IDmission customers continuously used to sharpen the models.

Delivered As

Facial Recognation

Mobile SDK

On-device liveness detection

ID Documents


Capture on web page and
verify against server



Liveness checks with single
frame or multiple frames

Pay per Use


Identity as a Service,
using just a QR code

Passive Liveness Facial Recognition Highlights


Machine learning technology

Trained extensively to detect spoofing attacks that are invisible to the naked eyes


Invisible by design

Completely passive. No need to blink, smile, turn head, or zoom in/out


Performance based on real world data

100,000 faces from Latin America, Africa, North America, South Asia and Asia Pacific


Measured False Acceptance Rate: 0%
Measured False Rejection Rate: 0%


Embed in your mobile app or run in the cloud


Average capture time < 2 Sec.