In computer security, biometrics relates to automatic authentication of individuals using measurable human characteristics, such as fingerprints, face recognition and iris recognition. Biometric authentication is used as a form of identification and access control at facilities across the commercial and government spectrum.

Face/Whole Body Biometrics

Face and whole body biometrics aim at recognizing people from the way their faces and/or bodies look and how they move. The significant advances that are continuously being made in the area are constantly faced with ever growing expectations, which lead to a continuous expansion of the research field. Face and whole body biometric authentication systems are now expected to work “in the wild” where (possibly extreme) viewing angles and environmental conditions can vary significantly between the probes and galleries. Such systems are also expected to behave in a way that is not biased towards or against a specific demographic group. Additionally, they are expected to be deployable in resource-constrained devices. These examples constitute a large part of VIMAL’s research work in face and whole body biometrics.

Presentation Attack Detection

Biometric authentication systems are vulnerable to presentation attacks, where a fake instrument can be presented to the system, instead of or on top of the bona fide presentation, for the purpose of obfuscating the real presented identity or impersonating a specific target identity. VIMAL has been developing both hardware, software, and hybrid approaches to address this challenge from different angles. Our work also has pushed the research frontier in this area in different directions, such as the timeless learning of newly discovered attack during models’ deployment, and the natural language explanation of the decisions of presentation attack detection models.

Related projects