As the number of users increases to hundreds of millions or billions levels, the requirements on performance of algorithms for biometric identification will increase exponentially. Existing identification technologies have been developed for security or surveillance applications mainly, which prefer relatively high recall rates but still require manual confirmation. While the internet-based financial application has much stricter requirements with regard to accuracy, recall rates, and retrieval speed, the accurate biometric identification technology at large scale needs to be further investigated and improved.
- On biometrics entity dataset with more than 100 million subjects, to achieve better performance than state-of-the-art.
- Patent submission and impactful research paper publication.