The Hummingbird Project represents a novel approach aimed at improving biometric analysis. Funded under the Human-Like Computing call, our approach seeks to identify the strategies used by humans when identifying one another, and to build these into computer algorithms in order to improve performance. With improved computer algorithms, we then seek to strengthen performance further through using multiple biometrics. Finally, we seek to identify the situations in which humans and (improved) computer algorithms agree and disagree in their identification decisions in order to determine the best way to address disagreement resolution.
This project is funded by the EPSRC (grant # EP/R030839/1) and will run from March 2018 to August 2019.
