A framework for secure selfie-based biometric authentication in the cloud

Abstract

Cloud-based selfie authentication has multiple advantages over on-device selfie authentication: Cloud-based authentication can support nomadic access from multiple devices including those not owned by the user, can leverage cheap and scalable utility computing, and can enable rapid innovation by allowing new matching algorithms to be continually deployed with no need to update the local device. This chapter presents a framework for a cloud-based selfie biometric authentication, which is termed Selfie-Biometrics-as-a-Service (SBaaS). By leveraging Platform-as-a-Service (PaaS) concepts, the framework is designed to enable independent software vendors to develop extensions and add-ons to a provider’s core application. In particular, the framework creates an innovative marketplace for biometric algorithms by providing a standard pre-built interface for the development and submission of new matching algorithms. When an authentication request is submitted, a criteria is used to select an appropriate matching algorithm. Every time a particular algorithm is selected, the corresponding developer is rendered a micropayment. Also presented in this chapter are solutions for preserving the confidentiality of biometrics stored in the cloud. This can be achieved through the use of biocryptosystems, which are secure biometric architectures involving the conversion of biometric features into secure signals that can be stored in the biometric database but are still useable for authentication. To provide a concrete example, a case study of a selfie-based ocular recognition system is disclosed, and detailed descriptions are provided of the user and developer interfaces.