Dr. Angelo Genovese representing Università degli Studi di Milano (UMIL), addressed a speech to the industry and academia about the most used methods in the literature for biometric information fusion. UMIL’s study is performed by analyzing and highlighting the approaches that can be applied in the context of ABC systems. Based on research, UMIL proposes privacy-compliant fusion algorithm to achieve the state-of-the-art performance by training using external biometric data (e.g., public datasets).
Using external biometric data in the tests will allow for the travellers’ sensitive data, such as the biometric data, not to be required. Results show that an increase in accuracy while using multibiometric fusion algorithms is feasible in ABC systems. Also privacy-compliant algorithms and technology-neutral framework increase the accuracy and feasibility in ABC systems as well. The audience found the presentation interesting. After the presentation, one of the attendees asked why we couldn’t use the biometric data captured in ABCs. However, during the discussion that took place it was clarified that biometric data captured in ABCs cannot be always used for testing the algorithms, since it is protected by privacy laws.
The conference was well organized and presented a good opportunity for communication between projects.