A Survey Of mobile face biometrics

Face biometrics have attracted significant attention as a technology for secure access to mobile devices. This is because almost all smartphones have RGB cameras suitable for capturing faces, and the required user interaction is acceptable given the popularity of 'selfies'. Most of the traditional methods for face biometrics may not be amenable to native execution on mobile hardware due to their limited memory and computing power. Consequently, a number of algorithms specifically designed or adapted to the mobile environment have been proposed for face biometrics. However, the state-of-the- art related to face biometrics in a mobile environment is not well known. This paper thoroughly and critically surveys face biometrics in terms of face detection and normalization, recognition, and anti-spoofing methods proposed for mobile devices. The overall aim is to improve understanding and discuss the advantages and limitations of the existing methods. Further, challenges and future research directions are identified for further research and development.