Biometric Authentication with a Timeless Learner


BATL's approach has been going in two parallel directions. One direction explores the benefits of combining novel and different sensing modalities to capture the liveliness of a biometric presentation. The other direction explores innovative approaches to model designs to address known and unknown presentation attacks. One of the main thrusts of BATL's approach is its \emph{timeless} nature, which means its continuous adaptability to unknown attacks in data acquired during operation.