PATRON: Exploring respiratory signal derived from non-contact face videos for face anti-spoofing
作者：
Highlights：
• We are the first one to use the respiratory pattern for 3D face mask anti-spoofing.
• Our respiration based method relies on a contradictory principle as that of rPPG.
• Appropriate ROI and temporal signal extraction are provided for the anti-spoofing.
• Our method provides a novel, compact and relevant features for face anti-spoofing.
• We outperform several well-known methods on publicly available challenging dataset.
摘要
•We are the first one to use the respiratory pattern for 3D face mask anti-spoofing.•Our respiration based method relies on a contradictory principle as that of rPPG.•Appropriate ROI and temporal signal extraction are provided for the anti-spoofing.•Our method provides a novel, compact and relevant features for face anti-spoofing.•We outperform several well-known methods on publicly available challenging dataset.
论文关键词：Face anti-spoofing,r-PPG,Face authentication systems,Respiratory signal,3D face mask
论文评审过程：Received 10 January 2021, Revised 6 July 2021, Accepted 5 September 2021, Available online 20 September 2021, Version of Record 21 September 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.115883
