Improving Trustworthiness of Identity using Biometrics, Computer Vision and Cryptography
Slides (pdf) Size: 3.86MB Abstract
Biometrics forms an important part of solution in combating crimes and frauds involving identity. They include synthetic identity, identity theft and account take-over. In this talk, I will discuss how biometrics and computer vision, including document validation technologies can be used together to establish high identity assurance, binding assurance, and authentication assurance. These aspects of assurance are defined by various documents, including but not limited to ISO/IEC 29115, NIST 800-63 (US), TDIF (Australia), eIDAS (EU), and GPG-45 (UK). In practice, I will argue why privacy-preserving biometrics must be used to support the above applications instead of using raw biometric templates or images. Two such techniques will be used as case studies, namely, cancellable biometrics and (key-binding) biometric cryptosystem.
Biography
Norman Poh currently provides technical leadership to accelerate Truststamp's growth in the area of privacy-preserving biometrics, improving its products and services which focus on identity management addressing KYC/AML issues. Previously, he held two data science roles in two distinctive sectors, namely, financial forecasting within an oil and gas industry and modelling, predicting and understanding disease progression from healthcare and health administrative data.
Currently, he is appointed by University of Malta as an Affiliate Associate Professor. Prior to the aforementioned data science roles, he held several academic roles at University of Surrey, from Senior Lecturer (Associate Professor); to principal investigator of a project funded by the Medical Research Council under the New Investigator Research Grant, leading the research on modelling chronic kidney disease.
He has written more 100 peer-reviewed publications, including five best paper awards (AVBPA’05, ICB’09, HSI 2010, ICPR 2010 and Pattern Recognition Journal 2006), and was awarded two personal fellowships sponsored by the Swiss National Science Foundation. He was named the Researcher of the Year by the university of Surrey in 2011. He received his Ph.D. degree in information fusion (machine learning) in 2006 from the Swiss Federal Institute of Technology Lausanne (EPFL), Switzerland.
He is a co-director of the Biometric Winter School. Previously, he served as an external adviser to the European reference network for critical infrastructure protection (ERNCIP); an academic reviewer to several universities; and an associate editor of the IET Biometrics Journal and the Biometric compendium. He has also reviewed for major IEEE/IAPR-organised conferences and journals in computer vision, biometrics, and security.
Norman Poh
Trust Stamp and AiiD Global
