Age & Identity Verification

• GDPR Certification from ACCS 2:2021 Technical Requirements for Data Protection and Privacy

Our approach has three-key aspects; we never store any data, all processing is performed locally on the device so we never send data over any network, and finally any data that is processed is automatically deleted afterwards.

Article 9 of GDPR (UK and EU): The data used to estimate age, cannot be used to uniquely identify a specific natural individual, and therefore cannot be considered special category data. We have been advised by legal experts that BIPA (Biometric Information Privacy Act) in the US and DPIA (Data Protection Impact Assessment) in Europe does not apply.

• We design, train and fine-tune all our own algorithms.
• All algorithms run locally on the hardware – quick, no internet required (ICU Pro & ICU Lite) 

We use our own algorithms to recognise faces and verify age. We do not use publicly available databases for training or tuning like many other systems, favouring our own extensive datasets especially chosen to reduce gender and skin tone bias. Our algorithms are not publicly available, which makes our system, and customer data more secure.

In the case of age estimation for example the model is trained using millions of images of faces with known ages. The algorithm analyses the features of the person and associates that with the known age. In doing so, the algorithm essentially learns what feature a subject typically has for a certain age. So, when a new image is presented, the algorithm analyses the features, and from what it has learnt, estimates an age based on what age it has associated with those features.

• Complete intelligent solution
• Local processing
• Quick outputs

Our edge solution gives customers an immediate response. Both age verification and facial recognition technology processes are performed locally giving a more reliable output and faster response. 

An internet connection is only required for initial set up and updates which are free for the first 12 months.

• Non-instrusive
• Instant real-time results
• Included as standard

Our range of biometric products, for age estimation and access control, uses single-image passive liveness detection. We have built, trained and tuned our own neural networks using an extensive data sample containing millions of images, to create an in-house, bespoke Presentation Attack Detection technique.

Using multiple parameters to analyse the full scene, we can determine a genuine subject. The same frame used for facial matching is also used for liveness checking, resulting in minimal processing. By not having a separate dedicated method for spoof, fraudsters have no information on how to fool the system, making it more secure. We have included anti-spoof technology as standard across the biometrics range. Many organisations use a third party to supply their spoof detection technology, for an additional fee, but our products have this included at no extra cost.