Multiple Face Analysis and Liveness Detection Using CNN

EasyChair Preprint 6547

Abstract

 In the fast-developing world, there is always a need for authentication for most systems. Face Recognition is one such system. Face recognition is the process of identifying or verifying a person from an image captured from either image capturing device or an individual frame in a video. A computer is not capable of processing such high-level processes by itself. For detecting and recognize faces, advanced concepts like deep learning can be used. Face Recognition is used in several fields such as login mechanism, unlock the device, User Authentication, etc. They can also play a pivotal role in recognition of multiple areas such as Student Attendance, Entry Cameras, etc. where the multiple persons may enter/present at a single instance of time. Also, there is a vulnerability for security breaches using still images of printed or digital images. The multiple face recognition and liveness detection models detect multiple faces in the image and recognize them based on the sample images provided and also verifies the liveness of identified person in the frame.

Keyphrases: Convolutional Neural Network, Multiple Face Recognition, liveness detection

@booklet{EasyChair:6547,
  author    = {Logeswari Saranya and K Umamaheswari},
  title     = {Multiple Face Analysis and Liveness Detection Using CNN},
  howpublished = {EasyChair Preprint 6547},
  year      = {EasyChair, 2021}}