AI-Powered Liveness Detection Prevents Identity Fraud

Innovatrics’ industry-leading facial liveness detection is based on deep neural networks to detect and prevent remote identity verification frauds. Verified by iBeta, it is compliant with ISO 30107-3 Level 2 Presentation Attack Detection (PAD) testing.

Industry-Proven Liveness Detection Technology

With a proven track record of biometric projects delivered globally to various government and enterprise customers, Innovatrics’ digital onboarding and liveness detection technologies support various use cases such as e-KYC, customer onboarding, bank account opening, SIM card registration, employee registration, driver registration, e-VISA issuance, digital certificate issuance, and more. 

Our Digital Onboarding Toolkit offers flexibility for integrators on how they integrate and deploy technology for face presentation attack detection. 

See e-KYC references
Telco SIM registration
Finance Bank account opening
System integrators Customer onboarding
Telco SIM registration
Finance Customer onboarding
Finance Bank account opening
System integrators Identity Management Platform
Digital Security Onboarding as a service
System integrators Customer onboarding
Telco Customer onboarding
System integrators Insurance registration
System integrators Visitor management

Protect Your Business Against Identity Fraud

Leveraging in-house R&D, we are constantly working to improve our algorithms to fight fraudulent attempts. Comprising several neural networks, our liveness detection is trained to detect various types of presentation attack vectors such as print attacks, dolls, 3D silicone masks, and display attacks.

Cooperating continuously with our clients throughout the project life cycle, we can provide recommendations for implementation and configuration, as well as a swift response to quickly changing attacks for our algorithms to meet their current needs. 

Constant Improvements in Presentation Attack Detection

Our liveness detection can be fine-tuned to meet the requirements of a specific use case – be its priority reinforced security, seamless user experience, or chosen type of deployment. 

  • Utilize high-performing liveness detection to detect various types of presentation attacks and prevent fraud
  • Rely on a technology partner that guarantees optimization and improvements throughout the whole life cycle of the project
presentation attack detection

Don’t Compromise Between User Experience and Security

Eliminating friction for users, Innovatrics liveness detection works with industry-leading speed and accuracy, resulting in higher completion rates of new customer onboardings. 

User Experience As a Priority

When developing the components of Digital Onboarding Toolkit, we always have user experience in mind and aim to remove as much friction from the onboarding process as possible. 

  • Make it easy for users to take a selfie from which passive liveness check is performed, or ask them to perform a simple action such as a smile
  • Get liveness detection results with top-performing accuracy in milliseconds 
  • Increase completion rates with seamless passive liveness detection
Liveness Detection

Passive Liveness Detection

Built upon deep neural networks, our passive liveness check can distinguish a real face from an image with top-performing accuracy without any user interaction.

With only one frame needed to tell if the person is real and alive, the check is instantaneous and extremely difficult to spoof. 

iBeta Level 2-Compliant

Brisk Results

  • Running in the background of the face capture process, it only needs one frame and takes milliseconds to tell if the person is real and alive.

99,9 % Completion Rate

  • Being fast and seamless for users, our passive liveness detection results in virtually no abandonment during the onboarding process.


  • Utilizing the power of neural networks compressed into a lightweight mobile component, it can run directly on-device and can be used even in markets with poor or limited bandwidth capacity.
iBeta compliance
See documentation
passive liveness detection

Smile Liveness Detection

Our semi-passive liveness detection combines the security of active liveness detection where a user is required to perform some kind of action (respond to a challenge) with little or no user-experience trade-off. 

  • After capturing their selfie, users are simply asked to smile
  • Quicker capture time and seamless user experience increase completion rates
  • Recommended for applications where security is paramount or where legislation obliges the user to respond to a challenge
smile liveness detection

Active Liveness Detection

Our active liveness check requires users to perform the simple task of following a moving dot on a screen. Our algorithm monitors the movement of the pupils, evaluating whether the dot has been followed correctly.

  • When the dot starts moving on the screen, the user must follow this movement with their eyes
  • Recommended as an additional step after performing passive liveness to improve the security of the remote identity verification process
  • An easy challenge that works on any smartphone
See documentation
active liveness detection

In-House Technology

We are constantly working on improving our liveness algorithms to keep them capable of detecting regularly evolving spoof attacks and preventing fraud while making sure they are not ethnically, racially, or sexually biased.

Much attention is given to building and using our own datasets when training neural networks for presentation attack detection. Moreover, we have established thorough liveness testing carried out at our own R&D center to deliver the best performing and unbiased algorithms possible.

Learn more in our white paper
liveness detection

Tech Box

The proprietary technology developed by industry leader

Works on-device or on server-side

Works on web, Android, iOS



Any questions? Let’s talk!