Abstract
Biometric identification is biometric-based authentication on mobile devices that nowadays has become ubiquitous, especially in unattended (e.g., access control for banks) and consumer (e.g., mobile phone unlocking) applications. While, face, fingerprint and inherent behavioral biometrics using inbuilt sensors such as accelerometer for smartphones person authentication, they have yet not achieved the desired or required level of efficiency, security and usability. This chapter presents an uncontrolled multibiometric smartphone framework utilizing multitrait selfie and behavioral biometrics. In particular, the presented system authenticates subject by ocular and face selfie features. This new multimodal biometric system also takes silently into account micro-movements of the phone, movements of the user’s finger on the touchscreen while user is capturing the multitrait selfie and entering passcode simultaneously in a split-screen mode of the smartphone. Addition of micro-movements behaviors enhances not only the performance but also robustness against noise and spoofing attacks. For this study, we collected a mobile multimodal dataset (MultiTouchMove) of touchstroke and phone-movement patterns in the wild from 95 subjects, which is made publicly available by the authors. Preliminary experimental analysis, using public MOBIO Face, VISOB ocular, and MultiTouchMove mobile datasets, on accuracy and usability shows promising results