Presentation Attack Detection Using Time-of-Flight-Based rPPG and Depth Features

With the rising importance of biometric authentication in numerous security applications it is critical to implement Presentation Attack Detection (PAD) techniques to protect against counterfeit biometrics. We evaluate a PAD-approach that uses 3D Time-of-Flight (ToF) cameras to perform heart rate measurements based on remote photoplethysmography (rPPG) for PAD. Compared to Red-Green-Blue (RGB)-camera based approaches, ToF technology provides advantages, including integration of 3D depth information, insensitivity to electronic displays and independence from ambient light and robustness against its influence. We introduce a Spectral pulse Peak prominence Ratio (SPR) and motion based metric and evaluate their applicability to PAD and present first promising results.