Marina Gavrilova, University of Calgary, Canada

Marina L. Gavrilova is a Full Professor in the Department of Computer Science, University of Calgary, a head of the Biometric Technologies Laboratory and a Board Member of ISPIA. Her publications include over 200 journal and conference papers, edited special issues, books and book chapters in the areas of image processing, pattern recognition, machine learning, biometric and online security. She has founded ICCSA – an international conference series with LNCS/IEEE, co-chaired a number of top international conferences, and is Founding Editor-in-Chief of LNCS Transactions on Computational Science Journal. Dr. Gavrilova is on the Editorial Boards of the Visual Computer, International Journal of Biometrics, and six other journals. She has given over 50 keynotes, invited lectures and tutorials at major scientific gatherings and industry research centers, including at Stanford University, SERIAS Center at Purdue, Microsoft Research USA, Oxford University UK, Samsung Research South Korea and others. Dr. Gavrilova currently serves as an Associate Editor for IEEE Access, IEEE Transactions on Computational Social Systems, the Visual Computer and the International Journal of Biometrics, and was appointed by the IEEE Biometric Council to serve on IEEE Transactions on Biometrics, Behavior, and Identity Science Committee.

Adaptive and Reliable Decision Making for Multi-Modal Biometric Systems

The area of biometric, without a doubt, have advanced to the forefront of an international effort to secure societies from both physical and cyber threats. This keynote provides an overview of the state-of-the-art in multi-modal data fusion and biometric system design, linking those advancements with real-world applications.

The rapid development of massive databases and image processing techniques has led over the past decade to the significant advancements in both fundamental biometric research and in a relevant commercial product development. Typical biometric applications include banking, border control, law enforcement, medicine, e-commerce, smart sensors and consumer electronics. A variety of issues related to biometric system performance and analysis has been addressed previously. A high number of biometric samples, data variability, data quality, data acquisition, types of fusion and system architectures have been shown to affect an individual biometric system’s performance. Addition of new types of behavioral data, based on social interactions, presents unique challenges and opportunities. This keynote reviews current trends related to design of adaptive and reliable multi-modal biometric systems, with the focus on issues of security and privacy of person data. It supports the theoretical developments with the practical examples on the use of multi-modal biometrics in industrial applications, including city planning, finance, medicine and situation awareness systems.