ID R&D launches SDK for on-device passive biometric liveness

Passive biometric liveness checks from ID R&D can now be performed on the user’s device, with the release of the IDLive Face Mobile SDK. The move gives ID R&D’s clients greater flexibility in authentication and onboarding implementations, and supports facial liveness on standalone kiosks and physical access systems, which the company says is a growth area.

IDLive Face on-device liveness is designed to suit scenarios where devices are already deployed with face biometrics, and internet access is limited or unavailable.

The server-side version of IDLive Face was launched in 2019. Both versions use a single frame for presentation attack detection (PAD) compliant with the ISO/IEC 30107-3 Levels 1 and 2.

IDLive Face Mobile runs on a smaller footprint than its server-side cousin, with similar performance, according to the announcement.

“We are completely committed to delivering a facial liveness detection product that meets the highest accuracy standards while keeping the user experience frictionless,” ID R&D President Alexey Khitrov says. “Offering an on-device option provides our partners the flexibility to address more opportunities in the rapidly expanding market for face verification.”

Payface uses ID R&D’s on-device facial liveness for fast point-of-sale transactions with systems that do not necessarily have reliable internet connections.

“The partnership with ID R&D allows us to bring the best in financial innovation and technology to Brazil. The combination of IDLive Face with Payface’s expertise in payment technology has resulted in a solution that is safer and faster, giving us the edge to win important deals throughout different markets,” said Eládio Isoppo, CEO at Payface. “We are very excited to be one of the companies making the payment process more accessible and quick for shoppers.”

Payface has also partnered with digital payments company Cielo to provide in-person credit card payments through facial recognition authentication, LABS reports.

The partnership between the Brazilian companies is beginning with a trial at a São Paulo drugstore, with plans to expand to other drugstores. Customers preregister their biometrics and credit card with the Payface app.

Cielo says roughly 18 percent of the 6.2 billion transactions it processed in 2020 were performed with contactless technologies like QR codes and NFC.

ID R&D has been moving steadily into biometrics at the network edge, including inking a partnership with Synaptics earlier this year.

Detecting liveness with computer vision

Computer vision has a long way to go to match the performance of human vision in many ways, writes ID R&D Machine Learning Engineer Alexey Grigoriev in a company post, but AI-based spoof detection systems trained on features like image borders and interference patterns are what enables biometric PAD.

What Grigoriev refers to as “classic” methods for PAD are dependent on effective feature engineering, he explains. Convolutional neural networks and backpropagation algorithms make it possible for the software to ‘learn’ the importance of features, but can also amplify bias. ID R&D invests significantly in reducing bias from its training data, he says, resulting in CNN-based liveness detectors which “are more suitable for real-world applications.”

Article Topics

accuracy  |  biometric liveness detection  |  biometrics at the edge  |  facial recognition  |  ID R&D  |  ISO/IEC 30107-3  |  onboarding  |  passive authentication  |  passive facial liveness  |  Payface  |  SDK  |  spoof detection