Multimodal Biometric Fusion Lab

Mobile devices rarely trust a single biometric sensor — Face ID falls back to a passcode, fingerprint sensors get paired with behavioral checks, and enterprise apps often combine two modalities before granting access. This simulator makes that fusion decision tangible: 400 genuine and 400 impostor authentication attempts are drawn from realistic fingerprint and face score distributions and rendered as a live 3D point cloud, and you choose how the two scores get combined — AND, OR, or a continuously weighted sum — while a translucent decision wall shows exactly which attempts pass. Live readouts track the combined False Accept Rate, False Reject Rate, overall accuracy and an approximate Equal Error Rate, so you can see directly how fusion strategy trades security against convenience.