National scale face identification platform for financial services with 110M+ faces database
Alexandr Khanin, Founder and CEO of VisionLabs, will present VisionLabs’ unique solution as a part of Sberbank ecosystem. The speaker together with the audience will do a deep dive into an implemented and ongoing huge scale project of National scale face identification platform for financial services with over 110 million faces in its database in order to understand challenges and technical specifics.
The foundation of this and many other projects is VisionLabs LUNA – a cross-platform object recognition software based on proprietary deep neural networks (DNN) inference framework. To build cost-effective solutions, VisionLabs use know-hows in DNN quantization and acceleration. In terms of accuracy, VisionLabs is recognized as a top three best in the world by National Institute of Standards and Technology's face recognition vendor test, and LFW by University of Massachusetts challenges.
VisionLabs LUNA has become the foundation of Sberbank's biometric identification platform with already existing database of 110M+ Customers. To enable operation at a national scale VisionLabs LUNA processes about 1000 simultaneous identification requests (1-to-110 million faces) every second with allowed latency of 1 second.
At the end of Mr Khanin’s speech, Vadim Kotov, Chief Technology Officer, Corporate Investment Banking, Sberbank, will speak about the approach to building the complex customer ecosystem of Sberbank and will make some examples of the most interesting investments connected to FinTech.