The serviceable and convenient nature of the Face Recognition System (FRS) makes it a preferred way for access control and authentication for a wide range of application areas, from biometric passport, surveillance system, health care, law enforcement, banking services to user verification in the smartphone. Most of the current day FRS have a number of open challenges such as weaker liveness detection, makeup attacks, morphing attacks and privacy issues. As the FRS do not actively query for the liveness of the subject and verify if the person is alive. Taking the advantage of the vulnerabilities in current day FRS, intruders can fool the FRS using the Presentation Attacks (PA) (a.k.a spoofing attacks). An attacker can mimic being an authentic user by presenting a spoof biometric data (e.g., printed photo, face videos, 3D face mask). Such an attack can be addressed by adding a layer of security to the Face Recognition System to detect them and these approaches are generally called Presentation Attack Detection. In this work, we propose Remote photoplethysmography based Presentation Attack Detection to distinguish presentation attacks (spoofing attempts) between the real face and 3D mask face videos. Remote photoplethysmography has been used to determine the liveness of a subject in PAD by biological signals such as pulse from the face videos. In this thesis, we propose a set of complementary features for making the PAD better against 3D face masks. We evaluate the performance of the proposed approach on two publicly available 3D mask datasets - 3DMAD and HKBVMarsV1+ using the standard protocols. The proposed approach outperforms the performance under similar protocols as against the state-of-the-art. Further, the thesis also investigates the use of proposed approach for cross dataset evaluation by training on one kind of 3D face masks and test on unseen data 3D mask types in an effort towards generalization.